Rural and Indigenous communities are at risk of losing cell service thanks to a 2019 law intended to strip US telecom networks of Chinese-made equipment. And while local companies were promised reimbursements as part of the “rip-and-replace” program, many of them have so far seen little of the funding, if any at all.

The federal push to block Chinese telephone and internet hardware has been years in the making, but gained substantial momentum during the Trump administration. In May 2019 an executive order barred American providers from purchasing telecom supplies manufactured by businesses within a “foreign adversary” nation. Industry and government officials have argued China might use products from companies like Huawei and ZTE to tap into US telecom infrastructure. Chinese company representatives have repeatedly pushed back on these claims and it remains unclear how substantiated these fears are.

[Related: 8.3 million places in the US still lack broadband internet access.]

As The Washington Post explained on Thursday, major network providers like Verizon and Sprint have long banned the use of Huawei and ZTE equipment. But for many smaller companies, Chinese products and software are the most cost-effective routes for maintaining their businesses.

Meanwhile, “rip-and-replace” program plans have remained in effect through President Biden’s administration—but little has been done to help smaller US companies handle the intensive transition efforts. In a letter to Congress on Thursday, FCC Chairwoman Jessica Rosenworcel explained an estimated 40 percent of local network operators currently cannot replace their existing Huawei and ZTE equipment without additional federal funding. Although $1.9 billion is currently appropriated, revised FCC estimates say another $3 billion is required to cover nationwide rip-and-replace costs.

Congress directed the FCC to begin a rip-and-replace program through the passage of the 2020 Secure and Trusted Communications Networks Act, but it wasn’t long before officials discovered the $3 billion shortfall. At the time, the FCC promised small businesses 39.5 percent reimbursements for their overhauls. Receiving that money subsequently triggered a completion deadline, but that remaining 61.5 percent of funding has yet to materialize for most providers. Last week, Sen. Maria Cantwell (D-WA) announced the Spectrum and National Security Act, which includes a framework to raise the additional $3 billion needed for program participants.

In her letter to Congress on Thursday, Rosenworcel said providers currently have between May 29, 2024, and February 4, 2025, to supposedly complete their transitions, depending on when they first received the partial funding. Rosenworcel added that at least 52 extensions have already been granted to businesses due in part to funding problems. Earlier this year, the FCC reported only 5 program participants had been able to fully complete their rip-and-replace plans.

It’s unclear how much of the US would be affected by the potential losses of coverage. To originally qualify for the reimbursement funding, a telecom company must provide coverage to under 2 million customers. The Washington Post cited qualified companies across much of the nation on Thursday, including Alaska, Colorado, Michigan, Missouri, New Mexico, Tennessee, Kansas, and Oklahoma. 

“The Commission stands ready to assist Congress in any efforts to fully fund the Reimbursement Program,” Rosenworcel said yesterday.

The post Many rural areas could soon lose cell service appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

The US Space Force located a tiny experimental satellite after it spent two-and-a–half decades missing in orbit. Hopefully, they’ll be able to keep an eye on it for good—unlike the last time.

The S73-7 Infra-Red Calibration Balloon (IRCB) was dead on arrival after ejecting from one of the Air Force’s largest Cold War orbital spy camera systems. Although it successfully departed the KH-9 Hexagon reconnaissance satellite about 500 miles above Earth in 1974, the S73-7 failed to inflate to its full 26-inch diameter. The malfunction prevented it from aiding ground based equipment triangulate remote sensing arrays and thus rendered it yet another hunk of space junk.

It wasn’t long afterwards that observers lost sight of the IRCB, only to once again locate the small satellite in early 1990s. And then, they managed to lose it again. Now, after another 25 years, the US Space Force’s 18th Space Defense Squadron rediscovered the experimental device.

The S73-7 satellite has been rediscovered after being untracked for 25 years. New TLEs for object 7244 started appearing on Apr 25. Congrats to whichever @18thSDS analyst made the identification. pic.twitter.com/YJOow5o4ND

— Jonathan McDowell (@planet4589) April 29, 2024

Confirmation came through a recent post on X from Jonathan McDowell, an astrophysicist at the Harvard-Smithsonian Center for Astrophysics, who offered his “congrats to whichever… analyst made the identification.”

So how does a satellite disappear for years on end not once, but twice? It’s actually much easier than you might think. As Gizmodo explained on May 1, over 27,000 objects are currently in orbit, most of which are spent rocket boosters. These, along with various satellites, don’t transmit any sort of identification back to Earth. Because of this, tracking systems must match a detected object to a satellite’s predictable orbital path in order to ID it.

[Related: Some space junk just got smacked by more space junk, complicating cleanup.]

If you possess relatively up-to-date radar data, and there aren’t many contenders in a similar orbit, then it usually isn’t hard to pinpoint satellites. But the more crowded an area, the more difficult it is for sensors to match, especially if you haven’t seen your target in a while—say, miniature Infra-Red Calibration Balloon from the 1970s.

It’s currently unclear what information exactly tipped off Space Force to matching their newly detected object with the S73-7, but regardless, that makes it at least trackable above everyone’s heads. In all that time, McDowell’s data indicates the balloon has only descended roughly 9 miles from its original 500 mile altitude, so it’ll be a while before it succumbs to gravity and burns up in the atmosphere. Accounting for everything in orbit may sometimes be taken for granted, but it’s a vital component of humanity’s increasing reliance on satellite arrays, as well as the overall future of space travel.

The post Space Force finds a dead Cold War-era satellite missing for 25 years appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

Billionaire LinkedIn co-founder Reid Hoffman has recently released a video ‘interview’ with his new digital avatar, Reid AI. Built on a custom GPT trained on two decades’ worth of Hoffman’s books, articles, speeches, interviews, and podcasts, Reid AI utilizes speech and video deepfake technology to create a digital clone capable of approximating its source subject’s mannerisms and conversational tone. For over 14 minutes, you can watch the two Hoffmans gaze lovingly and dead-eyed, respectively, into the tech industry’s uncanny navel. In doing so, viewers aren’t offered a “way to be better, to be more human,” as the real Hoffman argues—but a way towards a misguided, dangerous, unethical, and hollow future.

Many people might shudder at the idea of unleashing a talking, animated AI avatar of themselves into the world, but the tech utopian “city of yesterday” investor sounds absolutely jazzed about it. According to an April 24 blog post, he finds the whole prospect so “interesting and thought-provoking,” in fact, that he recently partnered with generative AI video company Hour One and the AI audio startup 11ElevenLabs to make it happen. (If that latter name sounds familiar, it’s because 11ElevenLabs’ product is what scammers misused to create those audio deepfake Biden robocalls earlier this year.)

After teasing a showcase of his digital clone for months, Hoffman finally revealed a (heavily edited) video conversation between himself and “Reid AI” last week. And what does the cutting-edge, deepfake-animated culmination of a custom built GPT-4 chatbot reportedly trained on all things Hoffman? A solid question—and one that isn’t easy to answer after watching the surreal, awkward, and occasionally unhygienic simulated interaction.

“Why would I want to be interviewed by a digital version of myself?” Hoffman posits at the video’s outset. First and foremost, it’s apparently to summarize one of his books for an array of potential audience demographics: the smartest person in the world, 5-year-old children, Seinfeld fans, and Klingons. While Hoffman seems to love each subsequent Blitzscaling encapsulation (particularly the “smartest person” one) they all sound like it came from a ChatGPT prompt—which, technically, they did. The difference here is that, instead of only a text answer, the words get a Hoffman vocal approximation layered atop of a (still clearly artificial) video rendering of the man.

Amidst all his excitement, Hoffman—like so many influential tech industry figures—yet again betrays a fundamental misunderstanding of how generative AI works. Technology like OpenAI’s GPT, no matter how gussied up with visual and audio additions, is not capable of comprehension. When an AI responds, “Thank you” or “I think that’s a great point,” they don’t actually experience gratitude or think anything. Generative AI sees sentences as lines of code, each letter or space followed by the next, most probably letter or space. This can be adapted into conversational audio and dubbed to video personas, but that doesn’t change the underlying functionality. It simply received new symbolic input that influences what basically amounts to a superpowered autocorrect system. Even if its language is set to Klingon, as Reid AI offers at one point.

So when Reid AI warns Hoffman a wrong answer may result “because I misinterpreted the information you gave, or I don’t have the full context of your question,” Hoffman doesn’t pause to explain any of the above facts for viewers. He instead moves along to his next conversation point, which usually involves a plug for his books or LinkedIn.

[Related: A deepfake ‘Joe Biden’ robocall told voters to stay home for primary election.]

Meanwhile, Reid AI’s visual component is supposedly meant to simulate many of Hoffman’s conversational mannerisms and queues. Judging from Reid AI’s performance, these largely boil down to stilted attempts at “nodding vigorously,” “emphatically tapping to illustrate a point,” and “picking his nose.” As New Atlas points out, the moment at 10:44 is an odd quirk to include in such a clearly condensed and edited video—perhaps meant to illustrate some of humanity’s more awkward, relatable traits. If so, it does little to distract from the far more absurd and troubling sentiments said by both Hoffman’s.

Reid AI expounds on boilerplate techno-libertarian talking points for fostering a “framework that fuels innovation.” Hoffman repeatedly opines that any concerns about bias, privacy, labor, and digital ownership concerns are just “start[ing] with the negative and [not realizing] all the things that are positive.” The digital clone regurgitates bland, uncreative ways to spruce up Hoffman’s LinkedIn page, like adding “personal flair” such as a fun and colorful header image.

But the most worrisome moment arrives when Hoffman contends “Everyone should be asking themselves, ‘What can I do to help?’” make AI like digital avatars more commonplace. He even goes so far as to equate the current technological era to Europe’s adoption of the steam engine, which made it “such a dominant force in the entire world.” (Neither he, nor Reid AI, concede the other tools involved in the industrial revolution, of course—namely a colonialist system built on the labor of millions of exploited and enslaved populations.)

Hoffman says future iterations of Reid AI will add “to the range of capabilities, of things that I could do.” It’s an extremely telling sentiment—one implying people like Hoffman have no qualms with erasing any demarcation between their cloned and authentic selves. If nothing else, Hoffman has already found at least one task Reid AI can handle for him.

“I am curious to know what others’ thoughts are on how to mitigate impersonation and all other types of risks stemming from such a use-case and achieve ‘responsible AI,’” one LinkedIn user asked him in his announcement post’s comments.

“Great question… Here is Reid AI’s answer,” Hoffman responded alongside a link to a new avatar clip.

The post Watch a tech billionaire talk to his AI-generated clone appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

Boston Dynamics may have relocated the bipedal Atlas to a nice farm upstate, but the company continues to let everyone know its four-legged line of Spot robots have a lot of life left in them. And after years of obvious dog-bot comparisons, Spot’s makers finally went ahead and commissioned a full cartoon canine getup for its latest video showcase. Sparkles is here and like its fellow Boston Dynamics family, it’s perfectly capable of cutting a rug.

Unlike, say, a mini Spot programmed to aid disaster zone search-and-rescue efforts or explore difficult-to-reach areas in nuclear reactors, Sparkles appears designed purely to offer viewers some levity. According to Boston Dynamics, the shimmering, blue, Muppet-like covering is a “custom costume designed just for Spot to explore the intersections of robotics, art, and entertainment” in honor of International Dance Day. In the brief clip, Sparkles can be seen performing a routine alongside a more standardized mini Spot, sans any extra attire.

But Spot bots such as this duo aren’t always programmed to dance for humanity’s applause—their intricate movements highlight the complex software built to take advantage of the machine’s overall maneuverability, balance, and precision. In this case, Sparkles and its partner were trained using Choreographer, a dance-dedicated system made available by Boston Dynamics with entertainment and media industry customers in mind.

[Related: RIP Atlas, the world’s beefiest humanoid robot.]

With Choreographer, Spot owners don’t need a degree in robotics or engineering to get their machines to move in rhythm. Instead, they are able to select from “high-level instruction” options instead of needing to key in specific joint angle and torque parameters. Even if one of Boston Dynamics robots running Choreographer can’t quite pull off a user’s routine, it is coded to approximate the request as best as possible.

“If asked to do something physically impossible, or if faced with an environmental challenge like a slippery floor, Spot will find the possible motion most similar to what was requested and do that instead—analogously to what a human dancer would do,” the company explains.

Choreographer is behind some of Boston Dynamics’ most popular demo showcases, including those BTS dance-off and the “Uptown Funk” videos. It’s nice to see the robots’ moves are consistently improving—but maybe nice still is that it’s at least one more time people don’t need to think about a gun-toting dog bot. Or even what’s in store for humanity after that two-legged successor to Atlas finally hits the market.

The post Boston Dynamics gives Spot bot a furry makeover appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

The Federal Trade Commission has ordered Razer to issue over $1.1 million in full refunds for its Razer Zephyr facemasks after alleging the PC gaming accessory company falsely billed its futuristic “wearable air purifier” as equivalent to N95-grade respirators. In truth, the FTC says Zephyr’s makers never even submitted their product for testing to either the FDA or the National Institute for Occupational Safety and Health (NIOSH). 

Razer is best known for its sleek, futuristic, luminescent video gaming accessories—but during the height of COVID-19, the company specializing in RGB backlit keyboards and headphones thought it wise to wade into pandemic healthcare. Released in October 2021 following nearly a year of internet hype, the Razer Zephyr looked more like a cyberpunk cosplay accessory than an actual “wearable air purifier.” Still, the transparent, twin-fan mask included three replaceable filters supposedly functioned together as equivalents to existing N95-grade products.

Outlets approached the odd healthcare accessory with a mix of anticipation and skepticism after plans were revealed in January 2021, later considered the pandemic’s deadliest month in the US. In the months leading up to its official launch, Razer co-founder and CEO Min-Liang Tan repeatedly posted on social media “linking the mask to the rise of the COVID-19 Delta variant, making explicit health claims, positioning the mask as a reusable N95, and claiming that Razer was seeking certification… [but] knew that they had never sought—and were not seeking—such certification,” according to the FTC’s complaint.

[Related: Calling TurboTax ‘free’ is ‘deceptive advertising,’ says FTC.]

To qualify for N95 certification, filters must guard against at least 95-percent of ambient air particles between 0.1 and 0.3 micrometers in size, while also providing higher filtration rates for larger particulates. Although COVID-19 virus cells measure around just 0.1 micrometers or smaller, they are always bonded to larger bodies such as water molecules and other biological material, and thus are effectively blocked by N95-rated masks and filters.

Razer consulted with a Singapore-based quality assurance company during Zephyr’s development, and in emails wrote they intended to market the wearable as “N95 grade.” Subsequent reviews showed Razer’s design only achieved around 83 percent particulate filtration efficiency (PFE) while its fans were off, with just a three percent improvement with the fans enabled. Even then, FTC documents state the Razer Zephyr “frequently tested much lower” and “did not come close to consistently reaching a PFE of 95 percent.” The quality testing company even went so far as to warn against mentioning N95 ratings “as it is not relevant to this product, and the claim will cause confusion.” 

Despite this, Razer moved forward with its marketing and released Razer Zephyr in October 2021, amid spiking global COVID-19 rates due to the Delta variant. Masks and filter packs were made available online through limited drop releases, as well as at three physical locations in Seattle, San Francisco, and Las Vegas. A single mask and three sets of filter replacements retailed for $99.99, while a mask alongside 33 filter sets sold for $149.99. A single, 10-set filter pack cost its wearers $29.99. The company even announced plans for a “Pro” version featuring voice amplification in early January 2022.

Barely a week later, however, Razer began walking back its N95-grade marketing for Zephyr amid mounting scrutiny and criticism. The Pro edition never saw the light of day, and federal regulators eventually opened its official investigation into the situation. In addition to the more than $1.1 million in refunds, Razer must pay a $100,000 civil penalty, and is forbidden from making any future “COVID-related health misrepresentations or unsubstantiated health claims about protective health equipment.” All references to the sleek, shoddy masks now appear scrubbed from Razer’s official website.

“Products like the Zephyr invite a lot of scrutiny. Is this an honest, good-faith attempt to create an upgraded device for people who plan to wear masks in public long-term, or is it a cash grab? Does it work at all?” PopSci wrote in its official review from January 2022. “These are all good, fair questions to ask when a company with no history making medical technology quickly develops and launches an expensive piece of kit.”

The post Surprise! That futuristic COVID mask was even sketchier than we thought appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

If we want to establish a permanent human presence on the moon, we need more detailed maps than the existing options, some of which date back to the Apollo missions of 1960’s and 1970’s. After more than ten years of collaboration between more than 100 researchers working at the Chinese Academy of Sciences (CAS), the newest editions of lunar topography are rolling out for astronomers and space agencies around the world.

As highlighted recently by Nature, the Geologic Atlas of the Lunar Globe includes 12,341 craters, 81 basins, and 17 different rock types found across the moon’s surface, doubling previous map resolutions to a scale of 1:2,500,000.

[Related: Why do all these countries want to go to the moon right now?]

Although higher accuracy maps have been available for areas near Apollo mission landing sites, the US Geological Survey’s original lunar maps generally managed a 1:5,000,000 scale. Project co-lead and CAS geochemist Jianzhong Liu explained to Nature that “our knowledge of the Moon has advanced greatly, and those maps could no longer meet the needs for future lunar research and exploration.”

To guide lunar mapping into the 21st-century, CAS relied heavily on China’s ongoing lunar exploration programs, including the Chang’e-1 mission. Beginning in 2007, Chang’e-1’s high-powered cameras surveyed the moon’s surface from orbit for two years alongside an interference imaging spectrometer to identify various types of rock types. Additional data compiled by the Chang’e-3 (2013) and Chang’e-4 (2019) lunar landers subsequently helped hone those mapping endeavors. International projects like NASA’s Gravity Recovery and Interior Laboratory (GRAIL) and Lunar Reconnaissance Orbiter, as well as India’s Chandrayaan-1 probe all provided even more valuable topographical information.

The pivotal topographical milestone wasn’t an entirely altruistic undertaking, however. While CAS geophysicist Ross Mitchell described the maps as “a resource for the whole world,” he added that “contributing to lunar science is a profound way for China to assert its potential role as a scientific powerhouse in the decades to come.” 

[Related: Japan and NASA plan a historic lunar RV road trip together.]

The US is also far from the only ones anxious to set up shop on the moon—both China and Russia hope to arrive there by the mid-2030’s with the construction of an International Lunar Research Station near the moon’s south pole. Despite the two nations’ prior promise to be “open to all interested countries and international partners,” the US is distinctly not among the 10 other governments currently attached to the project.

China plans to launch its Chang’e-6 robotic spacecraft later this week, which will travel to the far side of the moon as the first of three new missions. In an interview on Monday, NASA Administrator Bill Nelson voiced his concerns of a potential real estate war on the moon.

“I think it’s not beyond the pale that China would suddenly say, ‘We are here. You stay out,’” Nelson told Yahoo Finance. “That would be very unfortunate—to take what has gone on on planet Earth for years, grabbing territory, and saying it’s mine and people fighting over it.”

But if nothing else, at least the new maps will soon be available to virtually everyone. The Geologic Atlas is included in a new book from CAS, Map Quadrangles of the Geologic Atlas of the Moon, which also features an additional 30 sector diagrams offering even closer looks at individual lunar regions. The entire map resource will soon also become available to international researchers online through a cloud platform called Digital Moon.

The post China compiled the most detailed moon atlas ever mapped appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

Online romance scams netted con artists over $1.1 billion in 2023, with an average reported loss of around $2,000 per target. These victims who span age, gender, and demographics often aren’t only out of money—their time, emotions, and sometimes even physical safety can be on the line. And while the CEO of the largest online dating company, Match Group, sympathizes, he contends that sometimes life just gives you lemons, apparently.

“Look, I mean, things happen in life. That’s really difficult,” Match Group CEO Bernard Kim told CBS Reports during an interview over the weekend, before adding they “have a tremendous amount of empathy for things that happen.”

“I mean, our job is to keep people safe on our platforms; that is top foremost, most important thing to us,” Kim continued. Kim’s statements come amid a yearlong CBS investigation series on online romance scammers, and the havoc they continue to inflict on victims. 

Match Group oversees some of the world’s most popular dating platforms, including Match.com, Tinder, Hinge, and OkCupid. According to its 2024 impact report, a combined 15.6 million people worldwide subscribe to at least one of its service’s premium features, with millions more utilizing free tiers. Although the FTC’s count of annual reported romance scams has declined slightly from its pandemic era highs, experts caution that these numbers could be vastly undercounted due to victims’ potential—and unwarranted—embarrassment.

Authorities believe as few as 7 percent of romance scams are actually reported, but while older age groups are frequently targeted, they aren’t alone. In fact, some studies show younger internet users are more likely to fall for online fraud than their elders, given a greater willingness to share personal information. Some of these con campaigns can span multiple years, and drain victims’ entire bank accounts and savings. At least one death has even been potentially tied to such situations.

[Related: Cryptocurrency scammers are mining dating sites for victims.]

Regulators and law enforcement agencies have attempted to hold companies like Match Group accountable as romance scam reports continue to skyrocket—an industry fueled in part thanks to the proliferation of tech-savvy approaches involving chatbots and other AI-based programs. In 2019, for example, the Federal Trade Commission filed a $844 million lawsuit alleging as many as 30 percent of Match.com’s profiles were opened for scamming purposes. A US District judge dismissed the FTC’s lawsuit in 2022, citing Match Group’s immunity against third-party content posted to their platforms.

Match Group says it invested over $125 million last year in its trust and safety strategies, and removes around 96 percent of new scam accounts created on any given day. The company reported a $652 million profit in 2023—up a massive 80 percent year-to-year.

[Related: Don’t fall for these online love scams.]

The FTC advises internet users to never send funds or any gifts to someone they never met in person, as well as keep trusted loved ones or friends informed of new online relations. Experts also caution against anyone who repeatedly claims they cannot meet in real life. Conducting reverse image searches of any photos provided by a new online acquaintance can help confirm fraudulent identities. The FTC also encourages anyone to report suspected frauds and scams here.

In its 2024 impact report, the company touted its inaugural “World Romance Scam Awareness Day” sponsored by Tinder alongside Mean Girls actor Jonathan Bennett, which promoted similar strategies. According to the event’s official website, however, the PSA event is technically called World Romance Scam Prevention Day.

The post Romance scams just ‘happen in life,’ says CEO of biggest dating app company in the US appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

Earlier this week, NASA’s Solar Dynamics Observatory (SDO) recorded a rarely seen event—four nearly-simultaneous flare eruptions involving three separate sunspots, as well as the magnetic filament between them. But as impressive as it is, the event could soon pose problems for some satellites and spacecraft orbiting Earth, as well as electronic systems here on the ground.

It may seem like a massive ball of fiery, thermonuclear chaos, but there’s actually a fairly predictable rhythm to the sun. Similar to Earth’s seasonal changes, the yellow dwarf star’s powerful electromagnetic fluctuations follow a roughly 11-year cycle of ebbs and flows. Although astronomers still aren’t quite sure why this happens, it’s certainly observable—and recent activity definitely indicates the sun is heading towards its next “solar maximum” later this year.

As Spaceweather.com notes, early Tuesday morning’s “complex quartet” of solar activity was what’s known as a “super-sympathetic flare,” in which multiple events occur at nearly the same time. This happens thanks to the often hard-to-detect magnetic loops spreading across the sun’s corona, which can create explosive chain reactions in the process. In this case, hundreds of thousands of miles separated the three individual flares, but they still erupted within minutes of each other. All-told, the super-sympathetic flare encompassed about a third of the sun’s total surface facing Earth.

[Related: Why our tumultuous sun was relatively quiet in the late 1600s]

And that “facing Earth” factor could present an issue. BGR explains “at least some” of the electromagnetic “debris” could be en route towards the planet in the form of a coronal mass ejection (CME). If so, those forces could result in colorful auroras around the Earth’s poles—as well as create potential tech woes for satellite arrays and orbiting spacecraft, not to mention blackouts across some radio and GPS systems. The effects, if there are any, are estimated to occur over the next day or so, but at least they’re predicted to only be temporary inconveniences.

Luckily, multi-flare situations like this week’s aren’t a regular occurrence—the last time something similar happened was back in 2010 in what became known as the Great Eruption.

[Related: Hold onto your satellites: The sun is about to get a lot stormier]

Still, these super-sympathetic flares serve as a solid reminder of just how much of our modern, electronically connected society is at the sun’s mercy. As recently as 2022, for example, a solar storm knocked around 40 Starlink satellites out of orbit. The risk of solar-induced problems will continue to rise as the skies grow increasingly crowded.

While many companies continue to construct redundancy programs and backup systems for these potential headaches, astronomers and physicists still can’t predict solar activity very accurately. More research and funding is needed to create early warning and forecasting programs.

This year alone has already seen at least two other solar activity events—and seeing as how we still haven’t passed the solar maximum, more impressive (and maybe damaging) activity is likely on the way.

The post Rare quadruple solar flare event captured by NASA appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

A research team’s “bionic eye” deciphered thousands of new words hidden within an ancient scroll carbonized during the eruption of Mount Vesuvius—and the new text points to the long-lost, potential final resting place of the philosopher Plato.

The 1,800-scroll collection, located in the estate now known as the “Villa of the Papyri,” was almost instantaneously incinerated during the historic Mount Vesuvius eruption in 79 CE, before being buried in layers of pumice and ash. The latest discovery is part of ongoing global efforts focused on the ancient Greek library believed to belong to Julius Caesar’s father-in-law.

Although rediscovered in 1792, the trove of text remained almost entirely inaccessible due to the carbonized parchment’s fragility and blackened writing. In recent years, however, contributors to projects like the Vesuvius Challenge have worked to finally reveal the charred artifacts’ potentially invaluable information. In February, the project’s organizers announced that a team successfully translated 2,000 characters within a scroll through a combination of machine learning software and computer vision programming. Now, a separate group’s “bionic eye” has uncovered even more.

[Related: 2,000 new characters from burnt-up ancient Greek scroll deciphered with AI.]

According to Italian news outlet, ANSA, experts utilized infrared hyperspectral imaging alongside a relatively new approach known as optical coherence tomography (OCT) to see through the carbonized material. OCT employs cross-sectional, high-resolution imagery most often used by optometrists to photograph the back of the eye. In this case, however, combining the two tools allowed researchers to bypass the layers of carbon to read a major portion of the scroll by detecting faint evidence of handwriting that human eyes can no longer see.

Now, it appears the team helped solve a major mystery within the history of philosophy—the location of Plato’s grave. After translating the section, it appears Plato was finally buried in a garden near a shrine to the Muses at the Platonic Academy in Athens. What’s more, the text details the pivotal philosopher’s last night before reportedly succumbing to illness. Plato, suffering from a high fever, unfortunately wasn’t a fan of a nearby musician’s attempt to comfort him by playing “sweet notes” on flute. According to the scroll, he even went so far as to criticize their “scant sense of rhythm.”

The section also offers a revised timeline of Plato’s life by stating that the philosopher was sold into slavery in either 404 or 399 BCE. Before the new discovery, historians believed he was enslaved in 387 BCE.

Researchers aren’t stopping here, either. As Interesting Engineering notes, the team will use their “bionic eye” for further scans through 2026, while the Vesuvius Challenge will pursue its own methods to discover even more insights into the scrolls.

The post A ‘bionic eye’ scan of an ancient, scorched scroll points to Plato’s long-lost gravesite appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

It’s a bold pitch for homeowners: What if you let a small tech startup’s crowdfunded AI surveillance system dispense vigilante justice for you?

A Slovenia-based company called OZ-IT recently announced PaintCam Eve, a line of autonomous property monitoring devices that will utilize motion detection and facial recognition to guard against supposed intruders. In the company’s zany promo video, a voiceover promises Eve will protect owners from burglars, unwanted animal guests, and any hapless passersby who fail to heed its “zero compliance, zero tolerance” warning.

The consequences for shrugging off Eve’s threats: Getting blasted with paintballs, or perhaps even teargas pellets.

“Experience ultimate peace of mind,” PaintCam’s website declares, as Eve will offer owners a “perfect fusion of video security and physical presence” thanks to its “unintrusive [sic] design that stands as a beacon of safety.”

And to the naysayers worried Eve could indiscriminately bombard a neighbor’s child with a bruising paintball volley, or accidentally hock riot control chemicals at an unsuspecting Amazon Prime delivery driver? Have no fear—the robot’s “EVA” AI system will leverage live video streaming to a user’s app, as well as employ facial recognition technology system that would allow designated people to pass by unscathed.

In the company’s promotional video, there appears to be a combination of automatic and manual screening capabilities. At one point, Eve is shown issuing a verbal warning to an intruder, offering them a five-second countdown to leave its designated perimeter. When the stranger fails to comply, Eve automatically fires a paintball at his chest. Later, a man watches from his PaintCam app’s livestream as his frantic daughter waves at Eve’s camera to spare her boyfriend, which her father allows.

“If an unknown face appears next to someone known—perhaps your daughter’s new boyfriend—PaintCam defers to your instructions,” reads a portion of product’s website.

Presumably, determining pre-authorized visitors would involve them allowing 3D facial scans to store in Eve’s system for future reference. (Because facial recognition AI has such an accurate track record devoid of racial bias.) At the very least, require owners to clear each unknown newcomer. Either way, the details are sparse on PaintCam’s website.

But as New Atlas points out, there aren’t exactly a bunch of detailed specs or price ranges available just yet, beyond the allure of suburban crowd control gadgetry. OZ-IT vows Eve will include all the smart home security basics like live monitoring, night vision, object tracking, movement detection, night vision, as well as video storage and playback capabilities.

There are apparently “Standard,” “Advanced,” and “Elite” versions of PaintCam Eve in the works. The basic tier only gets owners “smart security” and “app on/off” capabilities, while Eve+ also offers animal detection. Eve Pro apparently is the only one to include facial recognition, which implies the other two models could be a tad more… indiscriminate in their surveillance methodologies. It’s unclear how much extra you’ll need to shell out for the teargas tier, too.

PaintCam’s Kickstarter is set to go live on April 23. No word on release date for now, but whenever it arrives, Eve’s makers promise a “safer, more colorful future” for everyone. That’s certainly one way of describing it.

The post Startup pitches a paintball-armed, AI-powered home security camera appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

Andreas Mogensen returned to Earth in mid-March after a six-and-a-half month stint aboard the International Space Station. To mark his tenure as part of NASA’s Crew-7 mission, the Danish European Space Agency (ESA) astronaut has shared his souvenir from undock day—a guided video tour of the ISS.

It’s been a month now since I left the International @Space_Station.

One of the very last things that I did on undock day, was film a tour of the Space Station. It is as much a keepsake for me as it is a way for me to share the wonder of the International Space Station with you.… pic.twitter.com/oFR0VXR06A

— Andreas Mogensen (@Astro_Andreas) April 12, 2024

“It’s been a month now since I left the [ISS],” Mogensen posted to X early Friday morning. “… It is as much a keepsake for me as it is a way for me to share the wonder of the International Space Station with you. Whenever I will miss my time onboard ISS, and especially my crewmates, I will have this video to look at.”

Mogensen began his show-and-tell in the space station’s front end, above which a docked SpaceX Dragon craft awaited to take him home on March 12. On his left is the roughly 114-by-22-foot Columbus module—a science laboratory provided by the ESA back in 2008. Across from the lab is the smaller Japanese Experiment Module (JEM), nicknamed Kibō, which arrived not long after Columbus.

From there, Mogensen provides a first-person look at various other ISS facilities, including workstations, storage units, bathrooms, gym equipment, multiple docking nodes, and even the station kitchen. Of course, given the delicate environment, that module looks more like another lab than an actual place to cook meals—presumably because, well, no one is actually cooking anything up there.

But the most stunning area in the entire ISS is undoubtedly the cupola, which provides a 360-degree panoramic view of Earth, as well as a decent look at the space station’s overall size.

[Related: What a total eclipse looks like from the ISS.]

Speaking of which, Mogenen’s video also does a great job showcasing just how comparatively small the ISS really is, even after over 25 years of module and equipment additions. At 356-feet-long, it’s just one yard shy of the length of a football field, but any given module or transit space is only a few feet wide. Factor in the copious amounts of cargo, equipment, supplies, experiment materials, as well as the over 8-miles of cabling that wire its electrical systems, and it makes for pretty tight living conditions. Near the end of Mogensen’s tour, it only takes him a little over a minute to glide through most of the entire station back to his original starting point.

Of course, none of that undercuts one of humanity’s most monumental achievements in space exploration. Although the ISS is nearing the end of its tenure (it’s scheduled for decommission in 2031), Mogensen’s keepsake is a great document of what life is like aboard the habitat. But for those now looking for an even more detailed tour, there’s always NASA’s virtual walkthrough.

The post Let this astronaut show you around the International Space Station appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

Before astronauts leave Earth’s gravity for days, weeks, or even months at a time, they practice aboard NASA’s famous parabolic flights. During these intense rides in modified passenger jets, trainees experience a series of stomach-churning ups and downs as the aircraft’s steep up-and-down movements create zero-g environments. Recently, however, a robot received similar education as their human counterparts—potentially ahead of its own journeys to space.

A couple years back, eight students at ETH Zürich in Switzerland helped design the SpaceHopper. Engineered specifically to handle low-gravity environments like asteroids, the small, three-legged bot is meant to (you guessed it) hop across its surroundings. Using a neural network trained in simulations with deep reinforcement learning, SpaceHopper is built to jump, coast along by leveraging an asteroid’s low-gravity, then orient and stabilize itself mid-air before safely landing on the ground. From there, it repeats this process to efficiently span large distances.

But it’s one thing to design a machine that theoretically works in computer simulations—it’s another thing to build and test it in the real-world.

Sending SpaceHopper to the nearest asteroid isn’t exactly a cost-effective or simple way to conduct a trial run. But thanks to the European Space Agency and Novespace, a company specializing in zero-g plane rides, the robot could test out its moves in the next best thing.

Over the course of a recent 30 minute parabolic flight, researchers let SpaceHopper perform in a small enclosure aboard Novespace’s Airbus A310 for upwards of 30 zero-g simulations, each lasting between 20-25 seconds. In one experiment, handlers released the robot in the middle of the air once the plane hit zero gravity, then observed it resituate itself to specific orientations using only its leg movements. In a second test, the team programmed SpaceHopper to leap off the ground and reorient itself before gently colliding with a nearby safety net.

Because a parabolic flight creates completely zero-g environments, SpaceHopper actually made its debut in less gravity than it would on a hypothetical asteroid. Because of this, the robot couldn’t “land” as it would in a microgravity situation, but demonstrating its ability to orient and adjust in real-time was still a major step forward for researchers. 

[Related: NASA’s OSIRIS mission delivered asteroid samples to Earth.]

“Until that moment, we had no idea how well this would work, and what the robot would actually do,” SpaceHopper team member Fabio Bühler said in ETH Zürich’s recent highlight video. “That’s why we were so excited when we saw it worked. It was a massive weight off of our shoulders.”

SpaceHopper’s creators believe deploying their jumpy bot to an asteroid one day could help astronomers gain new insights into the universe’s history, as well as provide information into our solar system’s earliest eras. Additionally, many asteroids are filled with valuable rare earth metals—resources that could provide a huge benefit across numerous industries back at home.

The post Watch a tripod robot test its asteroid leaping skills appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

Starting this week, Americans looking to purchase a new internet line will encounter a familiar looking box. Today, Internet service providers (ISP) and mobile broadband carriers began showing potential customers a Federal Communication Commission (FCC) mandated “broadband consumer label” that clearly explains how much they will have to pay for services and the estimated internet speeds they should expect to receive. Modeled after the ubiquitous nutrition labels in grocery stores, the FCC hopes these new labels could cut down on opaque hidden service fees and provide needed transparency to consumers trying to save money on broadband access. 

Internet providers are required to show the label anytime a consumer tries to purchase home or mobile broadband, be that online or in-person. The label must include a plan’s total monthly cost and note if there is a discounted introductory rate. Provides will also have to clearly list other additional costs like modem or router rentals, government taxes, and early termination fees. ISPs will have to provide estimated download and upload speeds as well as the total amount of data included in monthly plans where applicable.

Most importantly, the labels force providers to clearly list any separate, previously hidden fees they may charge in addition to the base line price. Critics have long argued these opaque additional charges, sometimes referred to as “junk fees” are needlessly confusing and lead consumers to pay more than they initially expect. Though most of the information included in the broadband labels is already publicly accessible, the FCC believes these prominent and easy to parse labels will give consumers more power to confidently understand what they are paying for and potentially give them the confidence to shop around for better deals. 

ISPs found in violation of the new measures could face penalties from the Federal Communication Commission under Section 503 of the Communications Act. While most broadband providers have until April 10 to implement the labels, ISPs with fewer than 100,000 lines have until October 10th of this year to do so. Several large providers like T-Mobile, Verizon, and Google Fiber have already released their labels.  

Hidden fees make it harder for some Americans to access affordable high-speed broadband 

Despite being increasingly necessary to operate in a digital first economy, access to fast, reliable internet in the US remains far from ubiquitous. A recent report from the FCC claims 24 million US residents currently lack access to high-speed broadband. Though a slew of economic and geographical factors contribute to that unequal access, the Biden administration has argued confusing, difficulty to parse “junk fees” make the problem worse. The administration cited recent research claiming hidden fees tacked on to broadband plans jacked-up overall plan prices by around 20%. 

“Junk fees cost American families tens of billions of dollars each year and inhibit competition, hurting consumers, workers, small businesses, and entrepreneurs,” the White House wrote in a statement. 

Other studies suggest “bundling,” the practice of combining TV, internet and phone services together under one bill, similarly make it difficult for consumers to know how much they’re paying for internet services on their own. The broadband nutrition labels, which must be shown whenever a consumer wants to purchase a new line, could bring some clarity to that notoriously opaque market. Consumer advocacy groups like Free Press and Consumer Reports have championed the labels and pushed for their implementation since 2009.

“Consumers are all too familiar with broadband bills that bury junk fees and service terms in the fine print.” Free Press Policy Director Joshua Stager said in a statement. “People deserve to know what they’re paying for, and this label will help.

ISPs, on the other hand, haven’t welcomed the labels with open arms. Trade groups representing some of the nation’s largest broadband providers have spent years lobbying to gut or remove the label requirement altogether. More recently, these groups fought the FCC over requirements forcing them to make the label clearly visible on consumers’ monthly billing statements. 

Supporters hope broadband labels can mimic food industry success 

FCC Bureau Chief for Consumer and Government Affairs Alejandro Roark reportedly told CNN this week the agency “borrowed” the label idea from figures already seen adorned on food packaging throughout the US. The Federal Drug Administration officially mandated food labels in 1990 following the passage of the Nutrition Labeling and Education Act in order to standardize food nutrition reporting. Studies suggest those labels could drive improved public health outcomes.

The new labels are part of a broader set of efforts by the federal government to push back on onerous, non-transparent fees throughout the economy. Regulators recently released policy capping credit card late fees and have proposed penalizing concert ticket sellers and hotels that don’t clearly list the cost of their products up front to consumers. 

Most internet users in the US should expect to see the legal beginning this week. Consumers can file a complaint with the FCC if they believe they notice a company out of compliance.

The post Internet providers finally forced to reveal all hidden fees on ‘nutrition’ label appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

Best overall		Netgear Nighthawk Cable Modem Wi-Fi Router Combo		SEE IT	This modem-router combination takes the guesswork out of Spectrum compatibilty.
Best cyber secure		Gryphon AC3000		SEE IT	Hackers are no match for this protective router.
Best budget		TP-Link AC1750		SEE IT	Lots of expensive features are packed into this budget-friendly router.

If you work from home—and you’re a Charter Communications customer—you know there’s a huge difference between a bad router and a router that’s compatible with Spectrum internet. A bad router means long wait times talking to customer service, annoyingly slow load speeds, and a pixelated presence on work video calls. Choosing the best router for work, play, and just surfing the web makes life easier—and reduces the amount of low to no bandwidth-induced headaches you might get. 

• Best overall: Netgear Nighthawk Cable Modem Wi-Fi Router Combo
• Best for gaming: ASUS AX5700 Wi-Fi 6 Gaming Router
• Best for streaming: Linksys – Dual-Band AX5400 Wi-Fi 6 Router
• Best mesh: NETGEAR Orbi Whole Home Tri-band Mesh Wi-Fi 6 System
• Best cyber secure: Gryphon AC3000
• Best budget: TP-Link AC1750

How we chose the best routers for Spectrum

We know how important a strong Wi-Fi connection is to effortless, lag-free performance on the best gaming laptops and streaming high-quality lossless audio. I personally have made the mistake of not doing my research before buying a router—would not recommend spending a few days Wi-Fi-less and with my own thoughts. To prevent you from facing a similar situation, we looked to personal testing, peer recommendations, critical reviews, and user impressions to find the best Spectrum routers.

The best routers for Spectrum: Reviews & Recommendations

The best routers for Spectrum can handle everything, whether it’s a day full of video meetings, nights scrolling TikTok, intense gaming sessions, or streaming lengthy explainer videos on YouTube. These are our top picks.

Best overall: Netgear Nighthawk Cable Modem Wi-Fi Router Combo

SEE IT

Why it made the cut: You don’t have to worry about purchasing a separate modem with this trusted combination device.

Specs

• Data transfer rate: 1,900 Mbps
• Coverage: 1,800 square feet
• Plan speed compatibility: Up to 400 Mbps
• Parental controls: Yes
• Ports: Four 1-gigabit Ethernet ports; one USB 2.0 port

Pros

• Easy installation
• Independent channels for 2.4 GHz and 5.0 GHz
• Warrants fast, reliable internet. 

Cons 

• Slow return on investment
• No support for Wi-Fi 6

The Netgear Nighthawk proves that two is better than one: This router also features a built-in modem, meaning you don’t have to shop for a Spectrum-compatible modem. This router covers 1,800 square feet for up to 30 devices and is compatible with Spectrum speed plans up to 400 Mbps. It includes separate networks for 2.4 GHz and 5.0 GHz bands, meaning you can organize your devices based on band level. You can plug in gaming consoles and computers to Ethernet thanks to four one-gigabit Ethernet ports, and can share storage with a connected device using the USB port. The router supports Wi-Fi 4 and 5 but does not support Wi-Fi 6. It features a WPA2-PSK security protocol, which is secure enough to protect your home without an enterprise authentication server—simply provide an 8- to 62-character-long passphrase to encrypt your network. However, it will take some time to see a return on your investment with the Nighthawk. Multiple reviews note that owners’ internet speed and performance were improved after setup, making the purchase worth it. Find more options for the best Netgear routers here.

Best for gaming: ASUS AX5700 Wi-Fi 6 Gaming Router

SEE IT

Why it made the cut: Reduced lag and latency are a tap away thanks to the router’s game mode and support for Wi-Fi 6. 

Specs

• Data transfer rate: 5700 Mbps
• Coverage: 2,500 square feet
• Plan speed compatibility: 1 Gbps
• Parental controls: Yes
• Ports: two 3.2 USB ports; one 2.5 G WAN/LAN port; one WAN port; four Ethernet ports

Pros

• Fiber internet compatible
• Easy installation
• Plenty of LAN ports for devices if need be

Cons 

• Upright design can hinder placement 

We are no strangers to the ASUS AX5700—we’ve previously named it the best all-purpose gaming router. And it is compatible with Spectrum’s Internet Gig plan, which can handle up to 1 Gpbs, making it a perfect addition to this list. It’s Wi-Fi 6 compatible and includes MU-MIMO functionality for increased speed and security. Download the ASUS Router app to turn on the mobile game mode, which reduces lag and latency for back-to-back Victory Royales. AiMesh support allows you to bring seamless, interruption-free coverage to your home—and, considering the price of mesh routers, that makes this one a steal. Hackers will be KO’d thanks to free lifetime access to ASUS AiProtection Pro, which includes WPA3 protection—the latest Wi-Fi security protocol—powered by Trend Micro, a cyber security software company. However, you can only set the router vertically, which could pose a problem if you need a horizontal router. 

Best for streaming: Linksys – Dual-Band AX5400 Wi-Fi 6 Router

SEE IT

Why it made the cut: Between the wide coverage range, support for more than 30-plus devices, and splittable price, this router means the only thing your roommates will lag on is doing their dishes. 

Specs

• Data transfer rate: 5.4 Gbps
• Coverage: 2,800 square feet
• Plan speed compatibility: 1 Gbps
• Parental controls: Yes
• Ports: 4 Ethernet ports; 1 internet port; 1 USB port

Pros

• Easy setup
• Separate guest access
• Easy device connection thanks to a WPS button

Cons 

• Reviews note trouble connecting the router to the Linksys app

We’ve all had the problem of the internet slowing because it’s slogged down by everyone’s respective phone, laptop, and gaming system streaming at once. This problem is exacerbated if your roommates like Internet-of-Things smart devices that connect to the network. Expect the slog to stop with this dual-band router, which can connect to more than 30 devices. Additionally, this router is Wi-Fi 6 compatible, meaning it’s fast and futureproof. This router includes four Ethernet ports, one internet port to connect the router to the modem, and one USB port. Also, the router can create a guest access network so your roommates’ siblings or frequent callers don’t cause security problems or snag too much bandwidth when they add their binge-watching to what’s pulling down all the data. A WPS button makes device connection easy, and you can use your smartphone, tablet, or computer browser for quick set-up. The second biggest problem with this router is deciding who will take it at the end of the lease. The largest problem is connecting it to the Linksys app.

Best mesh: NETGEAR Orbi Whole Home Tri-band Mesh Wi-Fi 6 System

SEE IT

Why it made the cut: Up to 5,000 square feet of coverage and an included satellite extender means you can take meetings almost everywhere in your home without lag.

Specs

• Data transfer rate: 4.2 Gbps
• Coverage: 5,000 square feet
• Plan speed compatibility: 1 Gbps
• Parental controls: Yes
• Ports: 1 Ethernet WAN port; 3 Ethernet ports; 2 Ethernet ports on satellite extender

Pros

• No dead zones
• Wide coverage area
• Tri-band Wi-Fi

Cons 

• Must pay Netgear to service software issues
• Expensive
• App capabilities are only free for 30 days

We’ve all experienced the pain of being laggy and pixelated on a video call, either from weak signals or dead spots in the home. Say goodbye to all that with the Netgear Orbi, which covers 5,000 square feet and includes a satellite extender for more reach. This router, our pick for working from home, can support up to 40 devices and has a crazy-fast data transfer rate of 4.2 Gbps, plus support for Wi-Fi 6 for futureproofing. Unlike other routers, this one supports a tri-band frequency: 2.4 GHz and 5 GHz band for connecting your devices, and a separate 5 GHz band so the router and included satellite can communicate with each other. Netgear Armor Antivirus and data theft protection powered by Bitdefender means business secrets will stay secret. However, there is a catch: this capability, along with others in the Orbi app like parental controls, are only free for 30 days. Afterward, you’ll need a subscription. And you’ll have to pay Netgear for software troubleshooting—that can be a lot of money to spend after purchasing an already-expensive router.

Best cyber secure: Gryphon AC3000

SEE IT

Why it made the cut: This router does not use a web browser for configuration, making it less likely that a hacker will jack your info. 

Specs

• Data transfer rate: 3 Gbps 
• Coverage: 3,000 square feet
• Plan speed compatibility: 1 Gbps
• Parental controls: Yes
• Ports: 1 WAN port; 3 Ethernet ports with backhaul capabilities

Pros

• Easy installation with Gryphon app
• Free intrusion detection in the first year
• Responsive support team

Cons 

• No advanced customization options

You won’t need to worry about hackers and internet thieves with the Gryphon AC3000, which offers advanced security thanks to its Wi-Fi 6 functionality, vulnerability scans, and app configuration that prevents hackers from accessing your network via the web. It also secures connected devices and sends alerts when it detects vulnerabilities, weak passwords, or infected devices. The router has fantastic basic security and you also get one year of intrusion detection for free with purchase. Afterward, it’s $89 per year. It covers 3,000 square feet—an average 2-3 bedroom home—at lightning-fast speeds thanks to its 3 Gbps data transfer rate. If you purchase multiple Gryphon routers, you can create a mesh network for even stronger, wider coverage. The Gryphon Connect app includes robust parental controls, including content filtering, screen time, and usage monitoring. Its user interface is incredibly streamlined and simple. However, reviews note that there could be more advanced customization options, like more device categories and fields for categories that should be blocked.

Best budget: TP-Link AC1750

SEE IT

Why it made the cut: The TP-Link AC1750 features high-end capabilities at a fraction of the cost. 

Specs

• Data transfer rate: 1750 Mbps
• Coverage: 2,500 square feet
• Plan speed compatibility: 400 Mbps
• Parental controls: Yes
• Ports: 1 USB port; 1 WAN port; 4 Ethernet ports

Pros

• VPN Server
• Alexa compatibility
• Bandwidth prioritization (QoS)

Cons 

• Not compatible with Wi-Fi 6

The TP-Link AC1750 tops Amazon’s list of computer routers for a reason: it includes a VPN server, bandwidth prioritization, parental controls, Alexa compatibility, and can connect up to 50 devices—all for under $100. A 2,500-square-foot range means you can take video calls outside and bandwidth prioritization allows you to assign devices to either of the dual bands to prevent congestion and slow internet speeds. Connect it to Alexa to turn the guest Wi-Fi on or off using your voice, or use the TP-Link Tether app to set up and manage your network. WPA/WPA2 wireless encryption keeps the entire family protected from hackers, and parental controls protect your kids from looking up mature content on the internet. The only downside of this router? It’s not Wi-Fi 6 compatible, which means it could become obsolete as new Wi-Fi protocols come out.

What to consider when buying the best routers for Spectrum

Routers are not one-size-fits-all. Here is what you need to know when shopping for the best routers for Spectrum:

Modem vs. router

A modem connects you to a wide area network, or WAN—the internet that Spectrum provides. A router connects your devices to a local area network, or LAN—your own little piece of the internet that Spectrum provides. You can’t connect to a router without a modem. Spectrum customers are required to use an authorized modem, and the company provides a preconfigured Wi-Fi router for a monthly fee. However, that adds up, and purchasing your own can save money in the long run. If you buy your own router, you don’t have to give it back if you cancel your service—you can use it with your new internet provider.

Compatibility with Spectrum

Make sure that the router your purchase is compatible with Spectrum. Otherwise, you’ll have to return it. You can find a list of Spectrum-compliant routers on the company’s site. Although buying the correct modem is more important, it never hurts to double-check. Also, the router can only run as fast as your internet plan provides. For example, if you have a router that can only reach speeds of 300 Mbps and have an internet plan for 200 Mbps, you’ll only max out at 200 Mbps.

Wireless protocol

The wireless protocol, or Wi-Fi standard, determines your router’s throughput and range. You can determine the protocol by finding the number 802.11 and a certain letter combination:

• 11ax (Wi-Fi 6): Wi-Fi 6 is the newest standard and delivers speeds up to 10 Gbps. However, not all devices and internet connections support Wi-Fi 6. You may not need to upgrade just yet, but switching to Wi-Fi 6 will help futureproof your internet connection. This standard is the fastest.  
• 11ac (Wi-Fi 5): This standard appears on most routers, and supports speeds up to 3.5 Gbps. 
• 11n (Wi-Fi 4): This standard supports speeds up to 600 Mbps and was the first to allow both 2.4 GHz and 5 GHz frequencies. This standard, compared to Wi-Fi 5 and 6, is the slowest. 

Range and signal strength

Wi-Fi range is determined by the kind of router you’re using, the wireless protocol the router follows, and the space you’re in. Wi-Fi signals have a harder time permeating through concrete, compared to wood, stucco, and other building materials. For example, 2.4 GHz Wi-Fi routers can reach up to 150 feet indoors and 300 feet outdoors and should be used if you’re looking for long-distance Wi-Fi. Routers running on 5 GHz bands can reach around one-third of these distances since it uses narrower wavelengths—because of this, you should choose a 5 GHz router for speed but only if your devices can be located nearer to your internet access point. Newer routers operate on both 2.4 GHz and 5 GHz bands to reach greater distances and achieve maximum throughput for devices in closer proximity.

Budget

The best routers range from $50 up to $500. The best Wi-Fi extenders can help with spotty internet, but Consumer Reports recommends using a mesh router system that relies on multiple systems to spread strong Wi-Fi signals. However, a mesh router system can be pricey. On the flip side, Wi-Fi extenders are cheaper and can be a perfect solution if you’re looking for more reliable coverage in certain areas of your home. However, poor placement of your extender won’t help your Wi-Fi woes and can cause more connectivity problems if your Wi-Fi extender creates a separate network that your device has to switch between.

Extra features

Many newer routers are compatible with voice assistants like Siri and Alexa, have parental controls, and provide separate networks for guests. Built-in VPNs and routers with multi-user, multiple-input, and multiple-output technology—also known as MU-MIMO—can help with security and network speed, respectively. However, if you would not benefit from the bells and whistles—and the added costs associated with them—then there’s no shame in going for a basic router.

FAQs

Final thoughts on the best routers for Spectrum

• Best overall: Netgear Nighthawk Cable Modem Wi-Fi Router Combo
• Best for gaming: ASUS AX5700 Wi-Fi 6 Gaming Router
• Best for streaming: Linksys – Dual-Band AX5400 Wi-Fi 6 Router
• Best mesh: NETGEAR Orbi Whole Home Tri-band Mesh Wi-Fi 6 System
• Best cyber secure: Gryphon AC3000
• Best budget: TP-Link AC1750

Whether you decide on a router that helps you jumpstart your streaming career, keeps your new battlestation in sync with the rest of the squad, or feeds your OLED TV buttery-smooth 4K video, the best routers for Spectrum will be suited to your lifestyle and internet needs. It doesn’t matter if you go ham-handed on the features or keep it simple—the best routers will get you streaming, web surfing, online gaming, and working with ease.

Why trust us

Popular Science started writing about technology more than 150 years ago. There was no such thing as “gadget writing” when we published our first issue in 1872, but if there was, our mission to demystify the world of innovation for everyday readers means we would have been all over it. Here in the present, PopSci is fully committed to helping readers navigate the increasingly intimidating array of devices on the market right now.

Our writers and editors have combined decades of experience covering and reviewing consumer electronics. We each have our own obsessive specialties—from high-end audio to video games to cameras and beyond—but when we’re reviewing devices outside of our immediate wheelhouses, we do our best to seek out trustworthy voices and opinions to help guide people to the very best recommendations. We know we don’t know everything, but we’re excited to live through the analysis paralysis that internet shopping can spur so readers don’t have to.

The post The best routers for Spectrum of 2024 appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

New data indicates a once-in-a-generation eclipse is a pretty surefire way to convince people to finally log off the internet—at least for a few minutes. According to estimates from cloud-computing provider Cloudflare, yesterday’s online traffic dropped between 40-60 percent week-to-week within the April 8 eclipse’s path of totality. In aggregate terms for the US, “bytes delivered traffic dropped by 8 percent and request traffic by 12 percent as compared to the previous week” around 2:00pm EST.

According to NASA, yesterday’s path of totality included a roughly 110-mile-wide stretch of land as it passed across Mazatlán, Mexico, through 13 states within the continental US, and finally over Montreal, Canada. In America alone, an estimated 52 million people lived within the eclipse’s path of totality. And it certainly seems like a lot of them put down their phones and laptops to go outside and have a look.

[Related: What a total eclipse looks like from space.]

As The New York Times highlights, Vermont saw the largest mass log-off, with an estimated 60-percent drop in internet usage compared to the week prior. South Carolinians, meanwhile, appeared to be the least compelled to take a computer break, since their traffic only dipped by around four percent.

Interestingly, you can also glean a bit about weather conditions during the eclipse from taking a look at Cloudflare’s internet usage map of the US. While most of the states within the event’s trajectory showcase pretty sizable downturns, Texas only experienced a 15 percent reduction. But given a large part of the Lone Star State endured severe weather conditions, it’s likely many people remained inside—maybe even online to livestream the views of the eclipse elsewhere.

[Related: The full sensory experience of an eclipse totality, from inside a convertible in Texas.]

So what were people doing if they weren’t posting through the eclipse? Well, snapping photos of the moment is always pretty popular, while NASA oversaw multiple volunteer research projects.

Judging from Cloudflare’s data, it didn’t take long for people to log back online once the eclipse ended above them. Usage appeared to spike back to pretty standard levels almost exactly in time with the event’s ending in any given state. No doubt most people rushed to post their reactions, photos, and videos… but maybe yesterday will still serve as a nice reminder that there’s a lot more to see when you take a break and go outside for a bit.

The post Internet use dipped in the eclipse’s path of totality appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

It could have been the perfect crime, had they used better spray paint.

Recently, authorities have seized over 320 lbs worth of suspected smuggled gold during a cargo freight search at Hong Kong International Airport, according to yesterday’s customs announcement. Bound for Tokyo on March 27, investigators recovered the roughly $10.7 million haul from within two actual air compressors, the bureau’s largest ever in terms of overall gold value. But these weren’t goldbond bricks or stacks of doubloons stashed deep within the machinery—they were hunks of precious metal molded into the shapes of compressor parts, then camouflaged with silver-colored spray paint.

Customs agents first noticed something suspicious after running the 1,708 lbs pair of air compressors through a security X-ray late last month during a standard screening process. As Business Insider explains, similar air compressors are made from aluminum or iron, and usually intended for industrial and mining projects, as well as to fill divers’ gas cylinders.

Speaking with the South China Morning Post (SCMP) on Monday, the assistant superintendent of Hong Kong International Airport’s customs air cargo division said technicians removed the motor casing and found a rotor “wrapped in a cord wheel which was tied to tape.”

“It was not similar to a normal motor,” he added.

After examining the rotor, authorities found traces of glue at both ends of the machinery part. Using a hammer, they then tapped the part and “noticed unevenness,” indicating the metal was far more malleable than it should have been. Scraping away at an outer layer of silver paint showed flecks of gold. At that point, the whole situation was pretty clear—these were dummy parts made of precious metal. Authorities believe the air compressor scheme was an attempt to evade Japan’s precious metals tariff that would have cost smugglers around $1.07 million, were they to go through official channels.

[Related: Montana traffickers illegally cloned Frankensheep hybrids for sport hunting.]

To create their industrial decoys, authorities believe smugglers must have first melted their gold down before pouring it into molds shaped to resemble motor rotors, screw shafts, and a gear piece. This probably was no easy feat, given that gold’s melting point is 1,948 degrees Fahrenheit.

According to Hong Kong Customs, police arrested the director of a local company on April 3 after finding his firm’s name listed as the shipment’s consignor. An initial investigation appears to show the company having no actual business dealings, potentially indicating it’s a shell outlet for smuggling. The investigation is still ongoing and the man has since been released on bail. Under Hong Kong’s Import and Export Ordinance, anyone found guilty of smuggling cargo could receive over $255,000 in fines alongside a maximum 7 year prison sentence.

The post Smugglers melted and spray painted $10 million in gold to look like machine parts appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

An international team of cryptographers has published a new whitepaper detailing the massive amounts of work, crowdsourcing, and computational programming that was required to translate a notorious serial killer’s half-century-old mystery message. Although one cryptographer uploaded a video rundown of their methodology to YouTube in 2020, the team’s new whitepaper further shows just how much work went into accomplishing their feat.

Between 1968-69, a man calling himself the Zodiac murdered at least five people in Northern California. During that time, as well as years after, the killer mailed a series of letters to local newspapers alongside a total of four ciphers. To this day, authorities have not formally named anyone as the Zodiac Killer, and only two of his cryptograms have been solved.

One of those, however, was long considered the most difficult to parse. First published in newspapers on November 12, 1969, the 340-character cipher (often referred to as “Z340”) baffled amateur and professional cryptographers alike for years. In December 2020, however, an international team announced they believed they finally cracked the Zodiac’s encoded message. A subsequent review by the FBI supported the solution offered by David Oranchak, Sam Blake, and Jarl Van Eycke, putting to rest a 51-year-old enigma.

[Related: Codebreakers have finally deciphered the lost letters of Mary, Queen of Scots.]

“I HOPE YOU ARE HAVING LOTS OF FUN IN TRYING TO CATCH ME,” the Zodiac Killer’s Z340 message begins, before clarifying he did not make the famous A.M. San Francisco television call-in on October 22, 1969.

THAT WASNT ME ON THE TV SHOW 

WHICH BRINGS UP A POINT ABOUT ME 

I AM NOT AFRAID OF THE GAS CHAMBER 

BECAUSE IT WILL SEND ME TO PARADICE ALL THE SOONER 

BECAUSE I NOW HAVE ENOUGH SLAVES TO WORK FOR ME 

WHERE EVERYONE ELSE HAS NOTHING WHEN THEY REACH PARADICE 

SO THEY ARE AFRAID OF DEATH 

I AM NOT AFRAID BECAUSE I KNOW THAT MY NEW LIFE IS

LIFE WILL BE AN EASY ONE IN PARADICE DEATH

Zodiac’s Z340 message, typos included

First spotted this week by 404 Media, the 39-page paper (accompanied by 23 pages of source materials) offers the fascinating and complex history behind Z340. According to the three authors, arriving at their ultimate solution had been preceded by “many years of failed experiments, dead-end ideas, and efforts to summarize what was known about the [Zodiac Killer].”

After countless fruitless attempts, the team felt confident that Z340 included some combination of homophonic substitution (one letter swapped for one or more symbols) and transposition (letters reordered according to a certain systematic logic). Unfortunately, that didn’t exactly narrow down the possibilities. As Discover Magazine explained in a 2021 profile, the cryptographers then faced hundreds of thousands of possible approaches to reading Z340.

To tackle all those potentials, the team turned to AZDecrypt, a program dedicated to homophonic decryption built by Van Eycke. The mathematical intricacies behind AZDecrypt are intense—but just for reference, the codebreakers say their program can solve up to 200 homophonic substitution ciphers per second with a 99 percent accuracy rate. After augmenting the software a bit to incorporate transposition options, AZDecrypt got to work, and soon yielded its first breakthroughs. Before long, the team finally unraveled Z340.

[Related: This ancient language puzzle was impossible to solve—until a PhD student cracked the code.]

Interestingly, the writers theorize it’s entirely possible the Zodiac Killer didn’t intend Z340 to be this difficult to decode. Speaking in 2021, Oranchak believes the computational power needed to ultimately break Z340 wasn’t even available in 1969. The Zodiac’s very first cipher, Z408, was decoded just days after being published, so it’s likely he meant to make Z340’s enciphering methods harder—but accidentally went too far as “a random unintended result of the encipherment process.” 

But as they make clear in their whitepaper, it wasn’t just computer software that solved one of the Zodiac Killer’s last mysteries. “The solution of this cipher was the result of a large, multi-decade group effort, and we ultimately stood on the shoulders of many others’ excellent cryptanalytic contributions,” they write.

The post How cryptographers finally cracked one of the Zodiac Killer’s hardest codes appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

The slide whistle isn’t known as a particularly difficult instrument to play—there’s a reason they’re usually marketed to children. But designing, programming, and building a robotic slide whistle quartet? That takes a solid background in computer science, a maddening amount of trial-and-error, logistical adjustments to account for “shrinkflation,” and at least two weeks to make it all happen.

That said, if you’re confident in your technical abilities, you too can construct a portable slide-whistle symphony-in-a-box capable of belting out Smash Mouth’s seminal, Billboard-topping masterpiece “All Star.” Fast forward to the 4:47 mark to listen to the tune. 

Despite his initial apology for “crimes against all things musical,” it seems as though Tim Alex Jacobs isn’t feeling too guilty about his ongoing robot slide whistle hobby. Also known online as “mitxela,” Jacobs has documented his DIY musical endeavors on his YouTube channel for years. It appears plans to create MIDI-controlled, automated slide whistle systems have been in the works since at least 2018, but it’s difficult to envision anything much more absurd than Jacob’s latest iteration, which manages to link four separate instruments alongside motorized fans and mechanical controls, all within a latchable carrying case.

Aside from the overall wonky tones that come from slide whistles in general, Jacobs notes just how difficult it would be to calibrate four of them. What’s more, each whistle’s dedicated fan motor differs slightly from one another, making the resultant pressures unpredictable. To compensate for this, Jacobs drilled holes in the pumps to create intentional air leaks, allowing him to run the motors closer to full power than before without overheating.

[Related: Check out some of the past year’s most innovative musical inventions.]

“If we can run them at a higher power level, then the effects of friction will be less significant,” Jacobs explains. But although this reportedly helped a bit, he admits the results were “far from adequate.” Attaching contact microphones to each slide whistle was also a possibility, but the work involved in calibrating them to properly isolate the whistle tones simply wasn’t worth it.

So what was worth the effort? Well, programming the whistles to play “All Star” in its entirety, of course. The four instruments are in no way tuned to one another, but honestly, it probably wouldn’t be as entertaining if they somehow possessed perfect pitch.

Jacobs appears to have plans for further fine tuning (so to speak) down the line, but it’s unclear if he’ll stick with Smash Mouth, or move onto another 90s pop-rock band.

The post Watch this robotic slide whistle quartet belt out Smash Mouth’s ‘All Star’ appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

Internet users in a handful of states across the US are finding it more difficult to browse parts of the web anonymously. Over a dozen states, including Texas and Louisiana, have enacted legislation forcing Pornhub and other purveyors of streaming online adult videos to verify the identities of its users to ensure children and teens aren’t accessing “sexual material harmful to minors.” Elsewhere, in states like Florida, lawmakers have introduced so-called online parental consent laws that would limit or ban underage users from accessing social media services over claims they cause psychological harm. In each case, lawmakers want online platforms to collect government-IDs from users or have them submit to third-party age verification methods to ensure they are indeed adults.

But determining whether or not kids and teens are actually accessing those sites means platforms have no choice but to verify the ages of all users accessing their sites, minor or otherwise. Adult porn viewers, who could previously dip in and out of websites with a relative degree of anonymity, may now fear having their government name and photograph at arms length away from their last Pornhub search query. At the same time, critics of the new laws worry some far-right, religiously conservative lawmakers could broadly interpret “adult” material to include content from LGBTQ+ creators or other people from marginalized groups who rely on the internet for a sense of community. In that scenario, teens from abusive or difficult family structures could find themselves shut out from support structures online. 

Experts speaking with PopSci say there are signs internet users in many of these states are turning to Virtual Private Networks (VPNs) to access otherwise blocked materials. Leading VPN provider Top10 VPN claims demand from VPN services jumped 275% on March 15, the same day Pornhub cut off access in Texas. The site says demand for VPNs similarly surged by 210% the day after a similar law took effect in Louisiana last year. ExpressVPN, another popular VPN provider, told PopSci it saw increased web traffic to its site the day anti-porn, online age verification bills took effect in seven out of eight states. 

“Wherever U.S. lawmakers have imposed age verification on internet users trying to access adult content online over the past 12 months, there has been a clear trend in the corresponding surges in demand for VPNs,” Top10 VPN Head of Research Simon Migliano told PopSci. In the most extreme case, Migliano claims Top10 VPN saw demand for the technology jolt up 847% the day the state’s new laws came into effect.

How are VPNs being used?

VPNs, which date back to the mid 1990s, create an encrypted tunnel for user’s data and can make it appear as if their computer is based in a different geographical location. Digital streaming viewers often use this VPN masking technology to access shows restricted in certain markets and blacked out sports events. Others view VPNs as useful tools for adding layers of security to private communications. That same technology has, for years, been used by whistleblowers, journalists, and political dissidents worldwide to bolster their anonymity online, especially in authoritarian countries. 

“A VPN is an effective tool for circumventing any kind of internet censorship, as it allows users to access the restricted content via an IP address from a location under a different jurisdiction from their own,” Migliano said. 

Though commonly used to bypass content restrictions in other countries, Center for Democracy and Technology Vice President of Policy Samir Jain says their apparent use by Americans to sidestep domestic content restrictions feels “relatively new.” That sudden shift, Jain said, owes itself partly to the language of these new laws which would have previously struggled to stand up to legal scrutiny. Jain, whose organization signed onto an amicus’s brief calling on a court to block the Texas law, said he wasn’t surprised users from affected areas states appeared to be seeking out VPNs. 

“If you provide a government ID to prove you are in effect no longer anonymous,” Center for Democracy and Technology Vice President of Policy Samir Jain told PopSci. “If people no longer feel like they can do that [access information anonymously] that infringes on their First Amendment expression right.”

ExpressVPN Privacy Advocate Lauren Hendry Parsons echoed that sentiment. 

“We know that when legislators restrict consumer access to services like porn, citizens still find a way to access it,” Hendry Parsons told PopSci. “There is absolutely a middle ground to be found that leans on third-party cooperation instead of limiting consumer rights.”

How are platforms responding to the new laws?

As of writing, seven mostly Republican-led states have passed some form of legislation relying on age-verification to prevent minors from accessing pornographic material. Nearly all of these so-called “age-gating” bills are copy-cat versions of a pioneering Louisiana legislation, which passed in 2022 and took effect early last year. The Verge estimates the Louisiana bill inspired at least 17 copycat bills, a handful of which are on their way to becoming laws. In Texas, sites found in violation of its law could face penalties of up to $10,000 per day. 

Some adult content sites like Pornhub have opted to block IP addresses originating from states with these new laws in order to avoid running afoul of the laws. Last month, internet users in Texas attempting to access the world’s largest purveyor of online adult video content were greeted instead with a 10 paragraph note from the company explaining its opposition to the state’s “ineffective, haphazard, and dangerous” law. Pornhub has similarly restricted access to users from half a dozen other states with similar age verification laws. In addition to wanting to steer clear of penalties, experts told PopSci platforms also oppose the laws because they don’t want to be responsible for collecting and maintaining torrents of sensitive users’ data that could pose a ripe target for cybercriminals.

“Age verification systems collect a huge amount of data, not only the personal information from each ID but also a record of each and every authentication made—essentially any site you access that features adult content,” Hendry Parsons said. “Combined with the data profiling social media companies create about their users, this treasure trove of personal information is a perfect target for bad actors.”

Rising VPN use could attract new lawmaker scrutiny

US internet users are reportedly using VPNs to access non-porn related material as well. College students around the country are reportedly already using VPNs to get around efforts from some universities to ban TikTok on campus networks. In Montana, where lawmakers passed a first-of-its kind statewide TikTok ban, creators have been preparing to similarly use the technology to stay connected to their followers. Lawmakers interested in restricting popular online content of various kinds will inevitably find themselves running into a VPN service willing to offer users an escape tunnel. 

But a continued uptick in VPN to access blocked risks inviting unintended consequences. Internet users appearing to use VPNs to blatantly run afoul of new legislation could incentivize lawmakers to clamp down on the technology. Some of the anti-porn laws, like the one enacted in Utah, already possess language explicitly prohibiting online platforms from letting minors “change or bypass restrictions on access.” Digital rights activists fear other recently proposed legislation aimed at limiting US user access to foreign apps may include provisions in it that would criminalize the use of VPNs.

Jain, from the Center for Democracy and Technology, acknowledged those concerns but said new laws banning criminalizing or restricting VPNs could do more harm than good and may face constitutional legal challenges. As for the new wave of laws appearing to fuel the rise in American VPN adoption, Jain said debates over one of those laws could eventually make its way up to the Supreme Court. 

“There are a lot of legitimate reasons to use VPNs to protect your privacy and anonymity,” Jain said.

The post Online porn restrictions are leading to a VPN boom appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

What time is it on the moon?

Well, right now, that’s somewhat a matter of interpretation. But humanity is going to need to get a lot more specific if it intends to permanently set up shop there. In preparation, NASA is aligning its clocks in preparation for the upcoming Artemis missions. On Tuesday, the White House issued a memo directing the agency to establish a Coordinated Lunar Time (LTC), which will help guide humanity’s potentially permanent presence on the moon. Like the internationally recognized Universal Time Zone (UTC), LTC will lack time zones, as well as a Daylight Savings Time.

It’s not quite a time zone like those on Earth, but an entire frame of time reference for the moon. 

As Einstein famously noted, time is very much relative. Most timekeeping on Earth is tied to Coordinated Universal Time (UTC), which relies on an international array of atomic clocks designed to determine the most precise time possible. This works just fine in relation to our planet’s gravitational forces, but thanks to physics, things are observed differently elsewhere in space, including on the moon.

“Due to general and special relativity, the length of a second defined on Earth will appear distorted to an observer under different gravitational conditions, or to an observer moving at a high relative velocity,” Arati Prabhakar, Assistant to the President for Science and Technology and Director at the Office of Science and Technology Policy (OSTB), explained in yesterday’s official memorandum. 

Because of this, an Earth-based clock seen by a lunar astronaut would appear to lose an average of 58.7 microseconds per Earth day, alongside various other periodic variational influences. This might not seem like much, but it would pose major issues for any future lunar spacecraft and satellites that necessitate extremely precise timekeeping, synchronization, and logistics.

[Related: How to photograph the eclipse, according to NASA.]

“A consistent definition of time among operators in space is critical to successful space situational awareness capabilities, navigation, and communications, all of which are foundational to enable interoperability across the U.S. government and with international partners,” Steve Welby, OTSP Deputy Director for National Security, said in Tuesday’s announcement.

NASA’s new task is about more than just literal timing—it’s symbolic, as well. Although the US aims to send the first humans back to the lunar surface since the 1970’s, it isn’t alone in the goal. As Reuters noted yesterday, China wants to put astronauts on the moon by 2030, while both Japan and India have successfully landed uncrewed spacecraft there in the past year. In moving forward to establish an international LTC, the US is making its lunar leadership plans known to everyone.

[Related: Why do all these countries want to go to the moon right now?]

But it’s going to take a lot of global discussions—and, yes, time—to solidify all the calculations needed to make LTC happen. In its memo, the White House acknowledged putting Coordinated Lunar Time into practice will need international agreements made with the help of “existing [timekeeping] standards bodies,” such as the United Nations International Telecommunications Union. They’ll also need to discuss matters with the 35 other countries who signed the Artemis Accords, a pact concerning international relations in space and on the moon. Things could also get tricky, given that Russia and China never agreed to those accords.

“Think of the atomic clocks at the US Naval Observatory. They’re the heartbeat of the nation, synchronizing everything,” Kevin Coggins, NASA’s space communications and navigation chief, told Reuters on Tuesday. “You’re going to want a heartbeat on the moon.”

NASA has until the end of 2026 to deliver its standardization plan to the White House. If all goes according to plan, there might be actual heartbeats on the moon by that point—the Artemis III crewed lunar mission is scheduled to launch “no earlier than September 2026.”

The post NASA is designing a time zone just for the moon appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

Each day an estimated 95 percent of the world’s data travels across the roughly 900,000 miles of submarine fiber optic cables criss-crossing the ocean floor. Modern life as we know it—from internet communications to video calls to streaming services—would look significantly different without this massive infrastructure. To keep up with the world’s insatiable data needs, construction could soon begin on a new cable located within a once-inaccessible environment.

Politico reports that a consortium of companies intends to move forward with the Far North Fiber project—a deep sea cable that would stretch over 9,000 miles through the Northwest Passage, connecting Europe to Japan, alongside additional landing sites in Alaska, Canada, Norway, Finland, and Ireland. Ironically, the potential endeavor is only possible due to one of the most pressing threats facing humanity.

As our digital lives travel along these submarine cables, they devour gigantic amounts of energy and further exacerbate climate change. The Arctic, for example, is currently warming almost four times faster than the rest of the planet, causing its sea ice to shrink by roughly 13 percent per decade. According to one Far North Fiber developer, however, all that terrifying environmental decimation creates a new business opportunity.

[Related: A 10-million-pound undersea cable just broke an internet speed record.]

The Arctic’s previously unthinkable thaws will present a “sweet spot where it’s now accessible and allows us a time window when we can get the cable safely installed,” Ik Icard, chief strategy officer at Far North Digital, told Politico.

Far North Fiber’s backers claim that, once constructed, their cable would also be better protected compared to similar lines elsewhere in the world. An estimated 100 to 150 lines are damaged every year globally, be it from accidental encounters with boat anchors and fishing equipment, or due to intentional sabotage.

The threat of sabotage is an increasing concern to the telecom companies overseeing deep sea cable systems. More than 90 percent of all Europe-Asia data traffic travels along cables within the Red Sea trading corridor. Thanks to a recent increase in the region’s geopolitical unrest and violence, cable lines face greater risk of damage. Just last month, three such lines were cut during ongoing Houthi rebel attacks on nearby shipping vessels.

Company representatives believe establishing a new route through the Northwest Passage could avoid similar issues in the future—at an estimated cost of €1 billion ($1.08 billion). That’s about four times the cost of laying a cable across the Atlantic Ocean, and around three times as much to do so in the Pacific. But despite the exponential price tag, the European Union has signaled its interest with a €23 million investment in Far North Fiber. The project’s developers also hope to convince the US and Canada to get involved. 

“Nobody wants to cut a cable under the ice, it’s really hard to do,” Far North Digital co-founder Ethan Berkowitz said.

A study published in Nature Reviews Earth & Environment estimates the Arctic could experience seasonally ice-free waters as soon as 2035—less than a decade removed from Far North Fiber’s proposed 2026 launch date.

The post Melting ice makes Arctic a target for a new deep sea internet cable appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

A completely free email service offering 1 GB of storage, integrated search capabilities, and automatic message threading? Too good to be true.

At least, that’s what many people thought 20 years ago today, when Google announced Gmail’s debut. To be fair, it’s easy to see why some AP News readers wrote letters claiming the outlet’s reporters had unwittingly fallen for Google’s latest April Fool’s Day prank. Given the state of email in 2004, the prospect of roughly 250-500 times greater storage capability than the likes of Yahoo! Mail and Hotmail sounded far-fetched enough—offering all that for free felt absurd.  But there was something else even more absurd than Gmail’s technological capabilities.

It’s hard to imagine now, but there was a time when forking over all your data to a private company in exchange for its product wasn’t the default practice. Gmail marked a major shift in strategy (and ethics) for Google—in order to take advantage of all those free, novel webmail features, new users first consented to letting the company vacuum up all their communications and associated data. This lucrative information would then be utilized to offer personalized advertising alongside sponsored ads embedded in the margins of Gmail’s browser.

“Depending on your take, Gmail is either too good to be true, or it’s the height of corporate arrogance, especially coming from a company whose house motto is ‘Don’t Be Evil,’” Slate tech journalist Paul Boutin wrote on April 15, 2004.

The stipulations buried within Gmail’s terms of use quickly earned the ire of watchdogs. Within a week of its announcement (and subsequent confirmation that it wasn’t an April Fool’s prank), tech critics and privacy advocates published a co-signed open letter to Google’s co-founders, Sergey Brin and Larry Page, urging them to reconsider Gmail’s underlying principles.

“Scanning personal communications in the way Google is proposing is letting the proverbial genie out of the bottle,” they cautioned. “Today, Google wants to make a profit from selling ads. But tomorrow, another company may have completely different ideas about how to use such an infrastructure and the data it captures.”

But the worries didn’t phase Google. Gmail’s features were truly unheard-of for the time, and a yearslong, invite-only rollout continued to build hype while establishing it as an ultra-exclusive service. The buzz was so strong that some people shelled out as much as $250 on eBay for invite codes.

As Engadget noted earlier today, Google would continue its ad-centric email scans for more than a decade. Gmail opened to the general public on Valentine’s Day, 2007; by 2012, its over 425 million active users officially made it the world’s most popular email service–and one of the most desirable online data vaults.

It would take another five years before Google finally acquiesced to intensified criticism, agreeing to end its ad-based email scanning tactics in 2017. By then, however, the damage was done—trading “free” services for personal data is basically the norm for Big Tech companies like Meta and Amazon. Not only that, but Google still manages to find plenty of ways to harvest data across its many other services—including allowing third-party app developers to pony up for peeks into Gmail inboxes. And with 1.5 billion active accounts these days, that’s a lot of very profitable information to possess.

In the meantime, Google’s ongoing push to shove AI into its product suite has opened an entirely new chapter in its long-running online privacy debate—one that began two decades ago with Gmail’s reveal. Although it debuted on April 1, 2004, Gmail’s joke is still on us all these years later.

The post Gmail debuted on April Fool’s Day 20 years ago. The joke is still on us. appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

In the average American house, any download rate above roughly 242 Mbs is considered a solidly speedy broadband internet connection. That’s pretty decent, but across the Atlantic, researchers at UK’s Aston University recently managed to coax about 1.2 million times that rate using a single fiber optic cable—a new record for specific wavelength bands.

As spotted earlier today by Gizmodo, the international team achieved a data transfer rate of 301 terabits, or 301,000,000 megabits per second by accessing new wavelength bands normally unreachable in existing optical fibers—the tiny, hollow glass strands that carry data through beams of light. According to Aston University’s recent profile, you can think of these different wavelength bands as different colors of light shooting through a (largely) standard cable.

[Related: No, ‘10G internet’ is not a thing.]

Commercially available fiber cabling utilizes what are known as C- and L-bands to transmit data. By constructing a device called an optical processor, however, researchers could access the never-before-used E- and S-bands.

“Over the last few years Aston University has been developing optical amplifiers that operate in the E-band, which sits adjacent to the C-band in the electromagnetic spectrum but is about three times wider,” Ian Phillips, the optical processor’s creator, said in a statement. “Before the development of our device, no one had been able to properly emulate the E-band channels in a controlled way.”

But in terms of new tech, the processor was basically it for the team’s experiment. “Broadly speaking, data was sent via an optical fiber like a home or office internet connection,” Phillips added. 

What’s particularly impressive and promising about the team’s achievement is that they didn’t need new, high-tech fiber optic lines to reach such blindingly fast speeds. Most existing optical cables have always technically been capable of reaching E- and S-bands, but lacked the equipment infrastructure to do so. With further refinement and scaling, internet providers could ramp up standard speeds without overhauling current fiber optic infrastructures.

[Related: An inside look at how fiber optic glass is made.]

“[It] makes greater use of the existing deployed fiber network, increasing its capacity to carry data and prolonging its useful life & commercial value,” said Wladek Forysiak, a professor at the Aston Institute of Photonic Technologies. In doing so, Forsyiak believes their solution may also offer a much greener solution to the world’s rapidly increasing data demands.

The post Researchers unlock fiber optic connection 1.2 million times faster than broadband appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

A website promising its AI service can accurately scan pictures of penises for signs of sexually transmitted infections is earning the ire of healthcare advocates and digital privacy experts, among many other critics. But while the internet (and Jimmy Fallon) have taken the makers of Calmara to task over the past week, it actually took two years to get here.

Where did the AI ‘intimacy bestie’ come from?

Back in 2022, the company HeHealth debuted itself as an online way to “get answers about your penis health in minutes.” To receive this information, the website uses a combination of questionnaires and what the company claims is a “65-96 percent accurate” AI screening tool allegedly trained on proprietary datasets to flag photographic evidence of various STIs, including genital warts, herpes eruptions, and syphilis. “Cancer” is also included in the list of scannable signs. If the results come back “positive”, HeHealth can then refer users to healthcare professionals for actual physical screenings, diagnoses, and treatment options. It’s largely flown under the radar since then, with only around 31,000 people reportedly using its allegedly anonymized, encrypted services over the last two years. And then came Calmara.

With a website overloaded with Gen Z-centric buzzwords, Calmara sells itself as women’s new “intimacy bestie,” offering to scan pictures of their potential sexual partners’ penises for indications of STIs. According to HeHealth CEO’s latest LinkedIn post, HeHealth and Calmara “are totally different products.” However, according to Calmara’s website, HeHealth’s owners are running Calmara, and it utilizes the same AI. Calmara also markets itself as (currently) free and “really in its element when focused on the D.”

In a March 19 reveal announcement, one “anonymous user” claimed Calmara is already “changing the conversation around sexual health.” Calmara certainly sparked a conversation over the last week—just not the one its makers likely intended.

A novelty app 

Both Calmara’s and HeHealth’s fine print concede their STI judgments “should not be used as substitutes for professional medical advice, diagnosis, treatment, or management of any disease or condition.” There’s an obvious reason why this is not actually a real medical diagnosis tool, despite its advertising. 

It doesn’t take an AI “so sharp you’d swear it aced its SATs” to remember that the majority of STIs are asymptomatic. In those cases, they definitely wouldn’t be visible in a photograph. What’s more, a preprint, typo-laden paper explaining Calmara’s AI indicates it was trained on an extremely limited image database that included “synthetic” photos of penises, i.e. computer-generated images. Meanwhile, determining its surprisingly accuracy is difficult to do—Calmara’s preprint paper says its AI is around 94.4-percent accurate, while the homepage says 95 percent. Scroll down a little further, and the FAQ section offers 65-to-90 percent reliability. Not a very encouraging approach to helping foster safe sex practices that would, presumably, require mutual, trustworthy statements about sexual health.

“On its face, the service is so misguided that it’s easy to dismiss it as satire,” sex and culture critic Ella Dawson wrote in a viral blog post last week. Calmara’s central conceit—that new intimate partners would be comfortable enough to snap genital photos for an AI service to “scan”—is hard to imagine actually playing out in real life. “… This is not how human beings interact with each other. This is not how to normalize conversations about sexual health. And this is not how to promote safer sex practices.”

No age verification

Given its specific targeting of younger demographics, Dawson told PopSci she believes “it’s easy to see how a minor could find Calmara in a moment of panic and use it to self-diagnose” which would constitute obvious legal issues, as well as ethical ones. For one, explicit images of minors could constitute sexual child abuse material, or CSAM. While Calmara expressly states its program shouldn’t be used by minors, it still lacks even the most basic of age verification protocols at the time of writing.  

“Calmara’s lack of any age verification, or even a checkbox asking users to confirm that they are eighteen years of age or older, is not just lazy, it’s irresponsible,” Dawson concludes.

Dubious privacy practices 

More to the point, simply slapping caveats across your “wellness” websites could amount to the “legal equivalent of magic pixie dust,” according to digital privacy expert Carey Lening’s rundown. While Calmara’s FAQ section is much vaguer on technical details, HeHealth’s FAQ page does state their services are HIPAA compliant because they utilize Amazon Web Services (AWS) “to collect, process, maintain, and store” data—which is technically true.

On its page dedicated to HIPAA regulations, AWS makes clear that there is no such thing as “HIPAA certification” for cloud service providers. Instead, AWS “aligns our HIPAA risk management program” to meet requirements “applicable to our operating model.” According to AWS, it utilizes “higher security standards that map to the HIPAA Security Rule” which enables “covered entities and their business associates” subject to HIPAA to use AWS for processing, maintaining, and storing protected health information. Basically, if you consent to use Calmara or HeHealth, you are consenting to AWS handling penis pictures—be them yours, or someone else’s.

[Related: A once-forgotten antibiotic could be a new weapon against drug-resistant infections.]

That said, Lening says Calmara’s makers may have failed to consider newer state laws, such as Washington’s My Health My Data Act, with its “extremely broad and expansive view of consumer health data” set to go into effect in late June. The first of its kind in the US, the My Health My Data Act is designed specifically to protect personal health data that may fall outside HIPAA qualifications. 

“In short, they didn’t do their legal due diligence,” Lening contends.

“What’s frustrating from the perspective of privacy advocates and practitioners is not that they were ‘embracing health innovation‘ and ‘making a difference‘, but rather that they took a characteristic ‘Move Fast, Break Things’ kind of approach to the problem,” she continues. “The simple fact is, the [online] outrage is entirely predictable, because the Calmara folks did not, in my opinion, adequately assess the risk of harm their app can cause.”

Keep Calmara and carry on

When asked about these issues directly, Calmara and HeHealth’s founders appeared nonplussed.

“Most of the criticism is based on wrong information and misinformation,” HeHealth CEO and Calmara co-founder Yudara Kularathne wrote to PopSci last Friday, pointing to an earlier LinkedIn statement about its privacy policies. Kularathne added that “concerns about potential for anonymized data to be re-identified” are being considered.

On Monday, Kularathne published another public LinkedIn post, claiming to be at work addressing, “Health data and Personally Identifiable Information (PHI) related issues,” “CSAM related issues,” “communication related issues,” and “synthetic data related issues.”

“We are addressing most of the concerns raised, and many changes have been implemented immediately,” Kularathne wrote.

When reached for additional details, Calmara CEO Mei-Ling Lu avoided addressing criticisms in email, and instead offered PopSci an audio file from “one of our female users” recounting how the nameless user and her partner employed HeHealth’s (and now Calmara’s) AI to help determine they had herpes.

“[W]hile they were about to start, she realized something ‘not right’ on her partner’s penis, but he said: ‘you know how much I sweat, this is heat bubbles,’” writes Lu. After noticing similar “heat bubbles… a few days later,” Stacy and her partner consulted HeHealth’s AI scanner, which flagged the uploaded photos and directed them to healthcare professionals who confirmed they both had herpes.

To be clear, medical organizations such as the Mayo Clinic freely offer concise, accurate information on herpes symptoms, which can include pain or itching alongside bumps or blisters around the genitals, anus or mouth, painful urination, and discharge from the urethra or vagina. Symptoms generally occur 2-12 days after infection, and although many people infected with the virus display either mild or no symptoms, they can still spread the disease to others. 

Meanwhile, Calmara’s glossy (NSFW) promotional, double entendre-laden video promises that it is “The PERFECT WEBSITE for HOOKING UP,” but no matter how many bananas are depicted, using AI to give penises a once-over doesn’t seem particularly reliable, enjoyable, or even natural.

The post Please think twice before letting AI scan your penis for STIs appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

SLIM, Japan’s first successful lunar lander, isn’t going down without a fight. After making history—albeit upside down—in January, the Smart Lander for Investigating Moon continues to surprise mission control at Japan Aerospace Exploration Agency (JAXA) by surviving not one, but now two brutally frigid lunar nights.

“Last night, we received a response from #SLIM, confirming that the spacecraft made it through the lunar night for the second time!” JAXA posted to X on Wednesday alongside a new image of its likely permanent, inverted vantage point near the Shioli crater. JAXA also noted that, because the sun is currently high above the lunar horizon, SLIM’s equipment is currently extremely hot (212-degrees Fahrenheit or so), so only the navigation camera can be used for the time being.

Last night, we received a response from #SLIM, confirming that the spacecraft made it through the lunar night for the second time! Since the sun was still high and the equipment was still hot, we only took some shots of the usual scenery with the navigation camera. #GoodAfterMoon pic.twitter.com/5BjIr7vxMG

— 小型月着陸実証機SLIM (@SLIM_JAXA) March 28, 2024

Based on their newly acquired data, however, it appears that some of the lander’s temperature sensors and unused battery cells are beginning to malfunction. Even so, JAXA says “the majority of functions that survived the first lunar night” are still going strong after yet another two-week stretch of darkness that sees temperatures drop to -208 Fahrenheit.

It’s been quite the multi-month journey for SLIM. After launching last September, SLIM eventually entered lunar orbit in early October, where it then spent weeks rotating around the moon’s surface. On January 19, JAXA initiated SLIM’s landing procedures, with early indications pointing towards a successful touchdown. After reviewing lander data, JAXA confirmed the spacecraft stuck the landing roughly 180-feet from an already extremely narrow 330-feet-wide target site—thus living up to SLIM’s “Moon Sniper” nickname.

[Related: SLIM lives! Japan’s upside-down lander is online after a brutal lunar night.]

The historic moment wasn’t a flawless mission, however. In the same update, JAXA explained that one of its lander’s main engines malfunctioned as it neared the surface, causing SLIM to tumble over, ostensibly on its head. In doing so, the craft’s solar panels now can’t work at their full potential, thus limiting battery life and making basic functions much more difficult for the lander.

JAXA still managed to make the most of its situation by using SLIM’s sensors to gather a ton of data on the surrounding lunar environment, as well as deploy a pair of tiny autonomous robots to survey the lunar landscape. On January 31, mission control released what it cautioned could very well be SLIM’s last postcard image from the moon ahead of an upcoming lunar night. The lander wasn’t designed for a lengthy life even in the best of circumstances, but its prospects appeared even dimmer given its accidental positioning.

Roughly two weeks later, however, SLIM proved it could endure in spite of the odds by booting back up and offering JAXA another opportunity to gather additional lunar information. A repeat of JAXA’s same warning came a few days later—and yet here things stand, with SLIM still chugging along. From the start, researchers have employed the lander’s multiple tools, including a Multi-Band Camera, to analyze the moon’s chemical composition, particularly the amounts of olivine, ““will help solve the mystery of the origin of the moon,” says JAXA.

At this point, it’s anyone’s guess how much longer the lander has in it. Perhaps it’s taking a cue from NASA’s only-recently-retired Mars Ingenuity rotocopter, which lasted around three years longer than intended.

The post Japan’s SLIM moon lander survives a second brutal lunar night appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

Somehow, someway, vinyl records keep defying the odds. Despite falling firmly behind compact disc sales for decades, the vintage physical music medium returned to the top spot in 2022 for the first time since 1987. Now, new numbers released by the Recording Industry Association of America (RIAA) indicate that wasn’t just a random fluke—yet again, vinyl outsold CDs for a second year running in 2023. This time, however, LPs managed to widen the lead even more.

As noted by The Verge on Tuesday, US music fans purchased around 43 million vinyl records in 2023, about 6 million more than total CD sales last year. What’s more, LP and EP purchases actually increased year-to-year by nearly 3 million sales—for the record (sorry), “long playing” vinyl are generally 12-inch records containing full albums usually played at 33 1⁄3 RPM, while “extended play” records are usually shorter, 7-inch releases spinning at 45 RPM. At the same time, people simultaneously bought less CDs this year than in 2022. All told, records generated nearly double the profit of their successor format—about $1.4 billion versus $437 million.

For reference, here’s a list of last year’s bestselling vinyl releases:

Unsurprisingly, it was a lot of Taylor Swift. But what’s more impressive to consider is that new vinyl is usually much more expensive to manufacture than compact discs.

[Related: Vinyl is back. But until now, record-making has been stuck in the ’80s.]

After spending decades as listeners’ go-to physical music medium, vinyl records finally passed the torch over to CDs back in 1987. For the next 35 years, that dynamic remained the same, with CDs’ comparative portability, durability, and overall audio quality making them the preferred method of enjoying music.

Sorry record purists, but it’s science. While transferring sound waves’ electrical signals into etched grooves on vinyl can offer “lossless” audio, that’s only under perfect conditions. And, as any vinyl owner knows, keeping a record in “perfect” condition is no easy task. This means that it doesn’t take much, or long, for an LP’s quality to degrade at least somewhat during playbacks.

[Related: This DVD-sized disk can store a massive 125,000 gigabytes of data.]

Contrast that to CDs, which digitally encode audio files onto optical discs that are translated back into sound via laser scanning. This can also sometimes result in diminished sound quality, but it’s overall a more reliable and standardized way to play music. Also (and perhaps most importantly) it’s generally a bit less of a hassle to keep CDs in good shape than vinyl upkeep, not to mention easier to travel with them.

But given the pros and cons of both options, it’s always really come down to how and where people want to enjoy their music—but for the second year in a row, it’s clear audiophiles are buying more vinyl instead of their once-unequivocal heirs.

The post Vinyl records outsold CDs for the second year running appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

It takes massive amounts of energy to power the data center brains of popular artificial intelligence models. That demand is only growing. In 2024, many of Silicon Valley’s largest tech giants and hoards of budding, well-funded startups have (very publically) aligned themselves with climate action–awash with PR about their sustainability goals, their carbon neutral pledges, and their promises to prioritize recycled materials. But as AI’s intensive energy demands become more apparent, it seems like many of those supposed green priorities could be jeopardized. 

A March International Energy Agency forecast estimates input-hungry AI models and cryptocurrency mining combined could cause data centers worldwide to double their energy use in just two years. Recent reports suggest tech leaders interested in staying relevant in the booming AI race may consider turning to old-fashioned, carbon-emitting energy sources to help meet that demand. 

AI models need more energy to power data centers 

Though precise figures measuring AI’s energy consumption remain a matter of debate, it’s increasingly clear complex data centers required to train and power those systems are energy-intensive. A recently released peer reviewed data analysis, energy demands from AI servers in 2027 could be on par with those of Argentina, the Netherlands, or Sweden combined. Production of new data centers isn’t slowing down either. Just last week, Washington Square Journal reports, Amazon Web Service Vice President of Engineering Bill Vass told an audience at an energy industry event in Texas he believes a new data center is being built every three days. Other energy industry leaders speaking at the event, like Former U.S. Energy Secretary Ernest Moniz, argued renewable energy production may fall short of what is  needed to power this projected data center growth. 

“We’re not going to build 100 gigawatts of new renewables in a few years,” Moniz said. The Obama-era energy secretary went on to say unmet energy demands brought on by AI, primarily via electricity, would require tapping into more natural gas and coal power plants. When it comes to meeting energy demands with renewables, he said, “you’re kind of stuck.” 

Others, like Dominion Energy CEO Robert Blue say the increased energy demand has led them to build out a new gas power plant while also trying to meet a 2050 net-zero goal. Other natural gas company executives speaking with the Journal, meanwhile claim tech firms building out data setters have expressed interest in using a natural gas energy source. 

Tech companies already have a checkered record on sustainability promises

A sudden reinterest in non-renewable energy sources to fuel an AI boom could contradict net zero carbon timelines and sustainability pledges made by major tech companies in recent years. Microsoft and Google, who are locked in a battle over quickly evolving generative AI tools like ChatGPT and Gemini, have both outlined plans to have net negative emissions in coming years. Apple, which reportedly shuttered its long-running car unit in order to devote resources towards AI, aims to become carbon neutral across its global supply chains by 2030. The Biden administration meanwhile has ambitiously pledged the US to have a carbon pollution free electricity sector by 2035.  

[ Related: Dozens of companies with ‘net-zero’ goals just got called out for greenwashing ]

Critics argue some of these climate pledges, particularly those heralded by large tech firms, may seem impressive on paper but have already fallen short in key areas. Multiple independent monitors in recent years have criticized large tech companies for allegedly failing to properly disclose their greenhouse gas emissions. Others have dinged tech firms for heavily basing their sustainability strategies around carbon offsets as opposed to potentially more effective solutions like reducing energy consumption. The alluring race for AI dominance risks stretching those already strained goals even further. 

AI boom has led to new data centers popping up around the US

Appetites for electricity are rising around the country. In Georgia, according to a recent Washington Post report, expected energy production within the state in the next ten years is 17 times larger than what it was recently. Northern Virginia, according to the same report, could require the energy equivalent of several nuclear power plants to meet the increased demand from planned data centers currently under construction. New data centers have popped up in both of those states in recent years. Lobbyists representing traditional coal and gas energy providers, the Post claims, are simultaneously urging government offices to delay retiring some fossil fuel plants in order to meet increasing energy demands. Data centers in the US alone were responsible for 4% of the county’s overall energy use in 2022 according to the IEA. That figure will only grow as more and more AI-focused facilities come online. 

At the same time, some of the AI industry’s-starkest proponents have argued these very same energy intensive models may prove instrumental in helping scale-up renewable energy sources and develop technologies to counteract the most destructive aspects of climate change. Previous reports argue powerful AI models could improve the efficiency of oils and gas facilities by improving underground mapping. AI simulation modes, similarly could help engineers develop optimal designs for wind or solar plants that could bring down their cost and increase their desirability as an energy source. Microsoft, who partners with OpenAI, is reportedly already using generative AI tools to try and streamline the regulatory approval process for nuclear reactors. Those future reactors, in theory, would then be used to generate the electricity needed to quench its AI models’ energy thirst. 

Fossil-fuel powered AI prioritizes long-term optimism over current day climate realities 

The problem with those more optimistic outlooks is that they remain, for the time being at least, mostly hypothetical and severely lacking in real-word data. AI models may increase the efficiency and affordability of renewable resources long term, but they risk doing so by pushing down on the accelerator of non-renewable resources right now. And with energy demands surging in other industries outside of tech at the same time, these optimistic longer-term outlooks could serve to justify splurging on natural gas and goal in the short term. Underpinning all of this is a worsening climate outlook that the overwhelming majority of climate scientists and international organizations agree demands radical action to reduce emissions as soon as possible. Renewable energy sources are on the rise in the US but tech firms looking for easier available sources of electricity to power their next AI projects risk setting back that progress. 

The post AI companies eye fossil fuels to meet booming energy demand appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

A team of miniature drones recently entered the radioactive ruins of one of Fukushima’s nuclear reactors in an attempt to help Japanese officials continue planning their decades’ long cleanup effort. But if the images released earlier this week didn’t fully underscore just how much work is still needed, new footage from the tiny robots’ excursion certainly highlights the many challenges ahead.

On Thursday, Tokyo Electric Power Company Holdings (TEPCO), the Japanese utility organization that oversees the Fukushima Daiichi plant reclamation project, revealed three-minutes of video recorded by a bread slice-sized flying drone alongside a snake-like bot that provided its light. Obtained during TEPCO’s two-day probe, the new clip offers viewers some of the best looks yet at what remains of portions of the Fukushima Daiichi nuclear facility—specifically, the main structural support in its No. 1 reactor’s primary containment vessel.

🎥 Fukushima: Vídeo de drone mostra interior do reator da usina nuclear danificado em 2011 pic.twitter.com/xs6HIGBLyV

— UOL Notícias (@UOLNoticias) March 22, 2024

The Fukushima plant suffered a catastrophic meltdown on March 11, 2011, after a magnitude 9.0 earthquake off the Japanese coast produced a 130-foot-tall tsunami that subsequently bore down on the region. Of the three reactors damaged during the disaster, No. 1 is considered the most severely impacted. A total of 880 tons of molten radioactive fuel debris is believed to remain within those reactors, with No.1 believed to contain the largest amount. An estimated 160,000 people were evacuated from the surrounding areas, with only limited returns allowed the following year. Around 20,000 people are believed to have been killed during the tsunami itself.

Last week’s drone-gathered images and video show the remains of the No. 1 reactor’s control-rod drive mechanism, alongside other equipment attached to the core, which indicate the parts were dislodged during the meltdown. According to NHK World, “agglomerated or icicle-shaped objects” seen in certain areas could be nuclear fuel debris composed of “a mixture of molten nuclear fuel and surrounding devices.”

[Related: Japan begins releasing treated Fukushima waste water into the Pacific Ocean.]

Experts say only a fraction of the damage could be accessed by the drones due to logistical difficulties, and that the robots couldn’t reach the core bottom because of poor visibility. Similarly, radiation levels could not be ascertained during this mission, since the drones did not include instruments such as dosimeters so as to remain light enough to maneuver through the plant.

TEPCO now plans to analyze the drone data to better establish a plan of action to collect and remove the radioactive debris within Fukushima. In August 2023, officials began a multiphase project to release treated radioactive wastewater from the plant into the Pacific Ocean. While deemed safe by multiple agencies and watchdogs, the ongoing endeavor has received strong pushback from neighboring countries, including China.

The Japanese government and TEPCO have previously estimated cleanup will take 30-40 years, although critics believe the timeline to be extremely optimistic.

The post Drones offer a glimpse inside Fukushima nuclear reactor 13 years after disaster appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

Got around 40 free weekends in the near future? Possess a 3D printer, extensive knowledge of vintage computer coding, soldering techniques, and near-superhuman patience? Then you, too, could be the proud owner of a “Brewintosh Plus,” a maddeningly accurate, completely working clone of Apple’s iconic Macintosh Plus computer system.

It might be hard to imagine, but there was a time when 1Mb of RAM was a big deal—and in 1986, the Mac Plus contained such immense processing power. To call Apple’s third Macintosh release a success is arguably an understatement. Until 2018, it remained the company’s longest-produced Macintosh model, with operating system updates made regularly until 1996.

It’s a pivotal piece of tech history, but finding one in decent condition, let alone complete working order, can be difficult after nearly four decades since its debut. For some collectors like Kevin Noki, however, the allure of tinkering with the iconic, retro hardware is too strong to resist. Unfortunately, it can be even harder to obtain a Mac Plus in places like Germany—where Noki happens to live.

But after scouring eBay for some time, Noki finally found and purchased an original, worse-for-wear 1Mb Macintosh Plus from eBay. Despite a broken power supply and missing floppy disk drive, one could technically emulate the original computer system simply by installing a Raspberry Pi and calling it a day—but that wouldn’t be much of a challenge, would it?

[Related: Macs are better at video gaming (emulators) than PCs. Here’s how to set up yours.]

Instead, Noki decided to use his vintage piece of tech history as a template for something much more accurate, if a bit more complicated: He built his own Mac Plus computer from the literal ground up.

“We are talking a properly sized, colored, and textured box, which takes wall power, swallows 3.5-inch disks, works with both telephone-cord and ADB Apple keyboards and mice, has a screen dimmer, and makes the startup sound (the beep, not the chord),” Ars Technica summarized earlier this week.

But even that laundry list of features doesn’t properly do Noki’s journey justice. At least 40 individual parts were measured, rendered into production specs through AutoDesk Fusion 460, and 3D-printed to create exact clones of the desktop’s many components. Then there was augmenting a USB floppy drive reader to use an Arduino-controlled motor that Noki coded himself, installing said floppy drive… not to mention soldering and wiring internal speakers, dyeing external parts to match the exact Mac Plus case color scheme, and even creating replicas of all its original labels, stickers, and raised-text stereotypes. But when it comes to picking the most difficult aspect of the entire saga, however, Noki doesn’t mince words.

“Honestly, everything was somewhat tough,” he tells PopSci, although he would wager that figuring out how to use an emulator to communicate with the rebuilt hardware was his biggest hurdle. “For instance, determining when to eject the floppy disk was particularly tricky, especially given my limited programming skills,” Noki says.

“Limited programming skills” is honestly pretty humble after watching Noki’s nearly hour-long YouTube rundown, which is genuinely worth an entire watch. Now that the job is done, the designer tells PopSci he’s gained an even greater respect for emulator programmers, “particularly the team responsible for the Mini vMac,” which simulates classic multiple Macintosh OS versions.

“Their dedication not only preserves computing history but also ensures its accessibility for generations to come, and for that, I’m incredibly thankful,” he says.

That thanks can certainly be extended to Noki, whose Brewintosh Plus and accompanying step-by-step guide now offers its own unique contribution to computing history preservation and accessibility.

The post A designer 3D printed a working clone of the iconic Mac Plus appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

Vernor Vinge, prolific science-fiction writer, professor, and one of the first prominent thinkers to conceptualize the concepts of a “Technological Singularity” and cyberspace, has died at the age of 79. News of his passing on March 20 was confirmed through a Facebook post from author and friend David Brin, citing complications from Parkinson’s Disease.

“Vernor enthralled millions with tales of plausible tomorrows, made all the more vivid by his polymath masteries of language, drama, characters, and the implications of science,” Brin writes.

The Hugo Award-winning author of sci-classics like A Fire Upon the Deep and Rainbow’s End, Vinge also taught mathematics and computer science at San Diego State University before retiring in 2000 to focus on his writing. In his famous 1983 op-ed, Vinge adapted the physics concept of a “singularity” to describe the moment in humanity’s technological progress marking “an intellectual transition as impenetrable as the knotted space-time at the center of a black hole” when “the world will pass far beyond our understanding.” The Singularity, Vinge hypothesized, would likely stem from the creation of artificial intelligence systems that surpassed humanity’s evolutionary capabilities. How life on Earth progressed from there was anyone’s guess—something plenty of Vinge-inspired writers have since attempted.

[Related: What happens if AI grows smarter than humans? The answer worries scientists.]

John Scalzi, bestselling sci-fi author of the Old Man’s War series, wrote in a blog post on Thursday that Vinge’s singularity theory in now so ubiquitous within science fiction and the tech industry that “it doesn’t feel like it has a progenitor, and that it just existed ambiently.”

“That’s a hell of a thing to have contributed to the world,” he continued.

In many ways, Vinge’s visions have arguably borne out almost to the exact year, as evidenced by the recent, rapid advances within an AI industry whose leaders are openly indebted to his work. In a 1993 essay further expounding on the Singularity concept, Vinge predicted that, “Within thirty years, we will have the technological means to create superhuman intelligence,” likening the moment to the “rise of human life on Earth.”

“Shortly after, the human era will be ended,” Vinge dramatically hypothesized at the time.
Many critics have since (often convincingly) argued that creating a true artificial general intelligence still remains out-of-reach, if not completely impossible. Even then, however, Vinge appeared perfectly capable of envisioning a dizzying, non-Singularity future—humanity may never square off against sentient AI, but it’s certainly already contending with “a glut of technical riches never properly absorbed.”

The post Vernor Vinge, influential sci-fi author who warned of AI ‘Singularity,’ has died appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

The first human patient to reportedly receive Neuralink’s wireless brain-computer interface (BCI) implant appeared to demonstrate the device’s early capabilities during a company livestream to X on Wednesday night. In late January, Elon Musk publicly stated that the experimental medical procedure was completed, but neither he nor his controversial medical startup had offered evidence of the results until yesterday evening’s 9-minute video.

Neuralink’s first volunteer is 29-year-old Noland Arbaugh from Texas, who dislocated his C4 and C5 vertebrae during a diving accident in 2016, permanently paralyzing him below the shoulders. During the livestream, Arbaugh appears to be playing online chess as a Neuralink BCI implant translates his brain activity into actionable computer inputs.

https://t.co/OMIeGGjYtG

— Neuralink (@neuralink) March 20, 2024

“If y’all can see the cursor moving around the screen, that’s all me,” he says at one point while highlighting a chess piece. “It’s pretty cool, huh?” According to Arbaugh, the key to successfully employing his new implant involved learning how to mentally differentiate between intentional and attempted movement—i.e, brain activity expressing the desire to move as opposed to activity which literally controls motor functions. “From there, I think it just became intuitive to me to start imagining the cursor moving,” Arbaugh continued, likening the feeling to “using the Force” from Star Wars.

Neuralink’s BCI is implanted using a robotic surgeon that subcutaneously connects the device’s microscopic wiring to a patient’s brain. Once installed, the hardware supposedly cannot be seen externally, and recharges wireless from “outside via a compact, inductive charger,” according to Neuralink’s website. Musk has repeatedly stated his hopes Neuralink will ultimately allow users to connect to the internet, smartphones, and computers through a line of upgradable, reversible BCI implants—and that he intends to one day receive the procedure himself.

“Long-term, it is possible to shunt the signals from the brain motor cortex past the damaged part of the spine to enable people to walk again and use their arms normally,” Musk claimed in a reply to Neuralink’s Wednesday evening post.

During the livestream hosted by Neuralink engineer Bliss Chapman, Arbaugh also described independently playing video games like the turn-based strategy game Civilization 6, which often entails more complex user inputs. Before the implant, Arbaugh says he frequently required assistance from his parents or a friend to play such games. Now, however, he says he has been able to do so for as long as eight hours and that the biggest impediment is simply waiting for the Neuralink’s battery to recharge.

[Related: Elon Musk alleges Neuralink completed its first human trial implant.]

“It’s not perfect. We have run into some issues,” Arbaugh concedes at one point, although he does not elaborate on the hurdles. “I don’t want people to think this is the end of the journey. There’s still a lot of work to be done.”

“We have more work to do. We have a lot to learn about the brain here,” agreed Chapman.

Neuralink is far from the first company to develop and install BCI implants, with the first successful commercial procedure dating back to 2010. Similar devices have since converted imagined handwriting into text, as well as thoughts into words. One competitor’s implant has also enabled users to browse the web as well as conduct online shopping and banking since 2019.

But it’s unclear if the reveal of Neuralink’s first human participant will assuage critics’ concerns regarding the company’s research record. Less than a day after Neuralink announced it would begin screening human volunteers for its multiyear ​​Precise Robotically Implanted Brain-Computer Interface (PRIME) Study last September, Wired released a damning exposé detailing graphic accounts of lab animal abuse during research. At the time, Wired’s coverage was one in a string of similar investigations into Musk’s company, including internal complaints of “hack job” surgical procedures resulting in over 1,500 animal deaths since 2018. The reports have since prompted multiple federal regulatory reviews and human trial delays.

As Wired also noted this week, Neuralink has not registered its PRIME Study on ClinicalTrials.gov, the federal documentation site for human medical studies, so information like how many human subjects Neuralink is seeking, where its procedures are taking place, or how its results will be assessed are not publicly available.

Still, Arbaugh encouraged people to consider applying to Neuralink’s ongoing PRIME Study, saying “there’s nothing to be afraid of” about the “super easy” procedure which he says has resulted in no cognitive impairments for him. Chapman, meanwhile, stated last night that additional updates on both Neuralink’s and Arbaugh’s progress will be released in the coming days.

The post Neuralink shows first human patient using brain implant to play online chess appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

Listening for crickets isn’t the only way you can help NASA conduct research during the total solar eclipse passing across much of North America on April 8—you can also lend your smartphone camera to the cause. The agency is calling on anyone within the upcoming eclipse’s path to totality to participate in its SunSketcher program. The program will amass volunteer researcher data to better understand the star’s shape. To participate, all you need is NASA’s free app, which uses a smartphone’s camera coupled with its GPS coordinates to record the eclipse. But why?

The sun looks simply spherical in many photographs and renderings, and in the sun if you happen to briefly glance at it during the day—an emphasis on “briefly,” of course. But thanks to what’s known as oblateness, this isn’t ever really the case. A rotating spheroid will oblate when its centrifugal force generates enough inertia to slightly flatten it out into a more irregular, elliptical shape. Within the solar system, Earth, Jupiter, and Saturn all also display oblateness, but the sun has some unique characteristics affecting how it oblates in particular.

According to NASA, the sun’s oblateness “depends upon the interior structure of the rotation, which we know from sunspot motions to be latitude-dependent at least.” Astronomers also think gas flows accompanying the sun’s magnetic activity and convection can create “transient distortions at a smaller level.” The upcoming total solar eclipse will provide astronomers an opportunity to better understand all this in the sun, but to make that happen, NASA wants you to harness the moon.

Earth’s natural satellite can serve as a valuable research partner in measuring the sun’s oblateness. This is due to a phenomenon known as “Baily’s beads,” which are the tiny flashes of light during an eclipse that occur as solar light passes over the moon’s rugged terrain of craters, hills, and valleys. Since satellite imagery has helped produce extremely detailed mappings of lunar topography, experts can match Baily’s beads to the moon’s features as it passes in front of the sun.

[Related: New evidence suggests dogs may ‘picture’ objects in their minds, similarly to people.]

These flashes will vary depending on where an observer is located within the path of totality. If you could amass data from a vast number of observer locales, however, you could better understand the sun’s surface variations due to its oblateness. And there are potentially millions of individual locales directly underneath the April 8 eclipse. Enter: SunSketcher.

“With your help, we hope to create a massive hour-long database of observations, more than we could ever make on our own,” NASA says.

All volunteers need to do is angle their phones up to capture the big event and let SunSketcher record the rest. Once all those videos are collected, NASA says the solar disk’s size and shape can be calculated to within a few kilometers, “an accuracy that is far better than currently known.” The reliable, detailed information on solar oblateness captured during SunSketcher can also be used to study how solar gravity affects the motions of inner planets, as well as help test various gravitational theories.

It’s worth noting that serving as an official SunSketcher volunteer will sacrifice the ability to use your smartphone to snap videos or pictures for yourself—but that’s arguably a small price to pay for helping conduct valuable scientific research.

The post NASA needs your smartphone during April’s solar eclipse appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

England’s court system has sentenced a “cyberflasher” to over a year in prison—a first for the country after its Online Safety Act went into effect on January 31. The 39-year-old culprit—already a registered sex offender—recently admitted in court to sending explicit photos of himself in February to both an adult woman and teenage girl via the messaging platform, WhatsApp. The woman subsequently took a screenshot of the interaction and reported it to police on the same day.

Passed by UK legislators last year, the new laws are designed to protect children and adults from exposure to unwanted imagery. They also place additional “legal responsibility on tech companies to prevent and rapidly remove illegal content, like terrorism and revenge pornography.”

A 2020 academic study determined roughly 76 percent of girls between 12-and-18 have received unsolicited sexual images from boys and men, often at random in the form of cyberflashing. The form of harassment is defined as sending unsolicited sexual imagery to targets via social media, text messages, or dating apps “for the purpose of their own sexual gratification or to cause the victim humiliation, alarm or distress,” and was added to the UK’s Online Safety Bill in March 2023 ahead of its formal passage the following October. Offenders can face a maximum of two years in prison if convicted.

[Related: How can you safely send nudes?]

“Just as those who commit indecent exposure in the physical world can expect to face the consequences, so too should offenders who commit their crimes online,” Hannah von Dadelzsen, Deputy Chief Crown Prosecutor for CPS East of England, said in an official statement on March 19.

As Engadget notes, similar digital legislative actions exist around the world, although they vary in scope and penalty. Scotland and Northern Ireland banned cyberflashing in 2010 and 2011, respectively, while both Australia and Singapore also enforce criminal charges for cyberflashing.

Here in the US, regulations continue on a more piecemeal basis. In 2022, California became the third state (after Texas and Virginia) to enact laws protecting against cyberflashing harassment. Dating app companies like Bumble have also voiced support for new laws to better prosecute cyberflashing. According to Bumble’s own internal surveying, despite billing itself as a “women-first” app, half of its women users have received such images on the platform. Attempts to address these issues at a federal level have yet to materialize in actual legislation.

Meanwhile, some lawmakers are attempting to leverage these legitimate concerns into wider-reaching censorship campaigns. In Oklahoma, for example, Republican state senators put forth a bill last month that seeks to ban exchanging all explicit content, even if solicited, for anyone except married couples as part of a broader anti-pornography push.

Following Tuesday’s conviction announcement, Deputy Chief Crown Prosecutor von Dadelzsen vowed “it will not be the last” of such prosecutions, and urged additional victims to come forward “knowing you have the right to lifelong anonymity” through England’s legal projections.

The post ‘Cyberflasher’ sent to prison for the first time in England appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

The medical startup Hippocratic AI and Nvidia have announced plans to deploy voice-based “AI healthcare agents.” In demonstration videos provided Monday, at-home patients are depicted conversing with animated human avatar chatbots on tablet and smartphone screens. Examples include a post-op appendectomy screening, as well as a chatbot instructing someone on how to inject penicillin. Hippocratic’s web page suggests providers could soon simply purchase its nursebots for less than $9/hour to handle such tasks, instead of paying an actual registered nurse $90/hour, Hippocratic claims. (The average pay for a registered nurse in the US is $38.74/hour, according to a 2022 U.S. Bureau of Labor Statistics’ occupational employment statistics survey.)

A patient’s trust in AI apparently is all about a program’s “seamless, personalized, and conversational” tone, said Munjal Shah, Hippocratic AI co-founder and CEO, in the company’s March 18 statement. Based on their internal research, people’s ability to “emotionally connect” with an AI healthcare agent reportedly increases “by 5-10% or more” for every half-second of conversational speed improvement, dubbed Hippocratic’s “empathy inference” engine. But quickly simulating all that worthwhile humanity requires a lot of computing power—hence Hippocratic’s investment in countless Nvidia H100 Tensor Core GPUs.

“Voice-based digital agents powered by generative AI can usher in an age of abundance in healthcare, but only if the technology responds to patients as a human would,” said Kimberly Powell, Nvidia’s VP of Healthcare, said on Monday. 

[Related: Will we ever be able to trust health advice from an AI?]

But an H100 GPU-fueled nurse-droid’s capacity to spew medical advice nearly as fast as an overworked healthcare worker is only as good as its accuracy and bedside manner. Hippocratic says it’s also got that covered, of course, and cites internal surveys and beta testing of over 5,500 nurses and doctors voicing overwhelming satisfaction with the AI as proof. When it comes to its ability to avoid AI’s (well documented) racial, gendered, and age-based biases, however, testing is apparently still underway. And in terms of where Hippocratic’s LLM derived its diagnostic and conversational information—well, that’s even vaguer than their mostly anonymous polled humans.

In the company’s white paper detailing Polaris, its “Safety-focused LLM Constellation Architecture for Healthcare,” Hippocratic AI researchers say their model is trained “on a massive collection of proprietary data including clinical care plans, healthcare regulatory documents, medical manuals, drug databases, and other high-quality medical reasoning documents.” And that’s about it for any info on that front. PopSci has reached out to Hippocratic for more specifics, as well as whether or not patient medical info will be used in future training.

In the meantime, it’s currently unclear when healthcare companies (or, say, Amazon, for that matter) can “augment their human staff” with “empathy inference” AI nurses, as Hippocratic advertises. The company did note it’s already working with over 40 “beta partners” to test AI healthcare agents on a wide gamut of responsibilities, including chronic care management, wellness coaching, health risk assessments, pre-op outreach, and post-discharge follow-ups.

It’s hard to envision a majority of people ever preferring to talk with uncanny chat avatars instead of trained, emotionally invested, properly compensated healthcare workers. But that’s not necessarily the point here. The global nursing shortage remains dire, with recent estimates pointing to a shortage of 15 million health workers by 2030. Instead of addressing the working conditions and wage concerns that led unions representing roughly 32,000 nurses to strike in 2023, Hippocratic claims its supposed cost-effective AI solution is the “only scalable way” to close the shortfall gap—a scalability reliant on Nvidia’s H100 GPU.

The H100 is what helped make Nvidia one of the world’s most lucrative, multi trillion-dollar companies, and the chips still support many large language model (LLM) AI supercomputer systems. That said, it’s now technically Nvidia’s third most-powerful offering, following last year’s GH200 Grass Hopper Super Chip, as well as yesterday’s simultaneous reveal of a forthcoming Blackwell B200 GPU. Still, at roughly $30,000-to-$40,000 per chip, the H100’s price tag is reserved for the sorts of projects valued at half-a-billion dollars–projects like Hippocratic AI.

But before jumping at the potential savings that an AI labor workaround could provide the healthcare industry, it’s worth considering these bots’ energy costs. For reference, a single H100 GPU requires as much power per day as the average American household.

The post Silicon Valley wants to deploy AI nursebots to handle your care appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

YouTube is flooded with fake livestream accounts airing looped videos of “Elon Musk” supposedly promoting crypto schemes. Although not the first time to happen, the website’s layout, verification qualifications, and search results page continue to make it difficult to separate legitimate sources from the con artists attempting to leverage today’s Starship test launch—its most successful to date, although ground control eventually lost contact with the rocket yet again.

After entering search queries such as “Starship Launch Livestream,” at least one supposed verified account within the top ten results takes users to a video of Elon Musk standing in front of the over 400-feet-tall rocket’s launchpad in Boca Chica, Texas. Multiple other accounts airing the same clip can be found further within the search results.

“Don’t miss your chance to change your financial life,” a voice similar to Musk’s tells attendees over footage of him attending a previous, actual Starship event. “This initiative symbolizes our commitment to making space exploration accessible to all, while also highlighting the potential of financial innovations represented by cryptocurrencies.”

“…to send either 0.1 Bitcoin or one Ethereum or Dogecoin to the specified address. After completing the transaction within a minute, twice as much Bitcoin or Ethereum will be returned to your address. …It is very important to use reliable and verified sources to scan the QR code and visit the promotion website. This will help avoid possible fraudulent schemes. Please remember administration is not responsible for loss due to not following the rules of our giveaway due to incorrect transactions or the use of unreliable sources. Don’t miss your chance to change your financial life. Connect Cryptocurrency wallet right now and become part of this amazing opportunity. You will receive double the amount reflected in your Bitcoin wallet. This initiative symbolizes our commitment to making space exploration accessible to all while also highlighting the potential of financial innovations are represented by cryptocurrencies. So let us embark on this remarkable journey to financial independence and cosmic discoveries…”

Fake Elon Musk

It’s unclear if the audio is AI vocal clone or simply a human impersonation, but either way oddly stilted and filled with glitches. A QR code displayed at the bottom of the screen (which PopSci cropped out of the video above) takes viewers to a website falsely advertising an “Official event from SpaceX Company” offering an “opportunity to take a share of 2,000 BTC,” among other massive cryptocurrency hauls.

There are currently multiple accounts mirroring the official SpaceX YouTube page airing simultaneous livestreams of the same scam clip. One of those accounts has been active since May 16, 2022, and has over 2.3 million subscribers—roughly one-third that of SpaceX’s actual, verified profile. Unlike the real company’s locale, however, the fake profile is listed as residing in Venezuela.

[Related: Another SpaceX Starship blew up.]

Scammers have long leveraged Musk’s public image for similar con campaigns. The SpaceX, Tesla, and X CEO is a longtime pusher of various cryptocurrency ventures, and is one of the world’s wealthiest men. Likewise, YouTube is a particularly popular venue for crypto grifters. In June 2020, for example, bad actors made away with $150,000 through nearly identical SpaceX YouTube channels. Almost exactly two years later, the BBC noted dozens of fake Musk videos advertising crypto scams, earning a public rebuke from the actual Musk himself. The crypto enthusiast outlet BitOK revealed a campaign almost exactly the same as today’s scams around the time as the November 2023 Starship event.

Update 3/15/24 12:40pm: YouTube spokesperson confirmed that the company has “terminated four channels in line with our policies which prohibit cryptocurrency phishing schemes.” According to YouTube, video uploads are monitored by a combination of machine learning and human reviewers.

The post Crypto scammers flooded YouTube with sham SpaceX Starship livestreams appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

What’s the weirdest thing you learned this week? Well, whatever it is, we promise you’ll have an even weirder answer if you listen to PopSci’s hit podcast. The Weirdest Thing I Learned This Week hits Apple, Spotify, YouTube, and everywhere else you listen to podcasts every-other Wednesday morning. It’s your new favorite source for the strangest science-adjacent facts, figures, and Wikipedia spirals the editors of Popular Science can muster. If you like the stories in this post, we guarantee you’ll love the show.

Heads up: Rachel and Jess are planning a livestream Q&A in the near future, as well as other fun bonus content! Follow Rachel on Patreon and Jess on Twitch to stay up to date. 

FACT: Dung beetles use the Milky Way to steer their poop balls 

By Christie Taylor

A dung beetle is a deceptively humble creature. Humble, in the sense that they literally eat and raise their young in poop. Deceptive in the sense that they perform incredible feats of strength, play an underappreciated ecological role, and can even (in some cases) appreciate our place in the cosmos.

Dung beetles are all members of the scarab family. And if the word “scarab” evokes ancient Egypt for you, there’s a reason. The scarab-headed god Khepri was, in fact, a dung beetle. Ancient Egyptians had seen these beetles rolling balls of dung, and connected them to the sacred, dung ball-like orb of the sun itself, as well as concepts of renewal and rebirth.

In any terrestrial ecosystem, a fresh pile of poop is an entire universe. Beetles eat dung, flies lay eggs, other insects come to eat the larvae of these animals. It’s a beautiful stinky circle of life.

It’s also dangerous and competitive, which is why one group of dung beetles has a strategy for success: get some poop, pack it into a ball, and roll it the hell out of there. These balls are sometimes 50 times their weight, and they’ll move them as far away as 200 meters before burying them underground.

First, though, they dance: the dung beetle hops atop their fresh-rolled ball of poop and performs one or more rotations before they then push their precious cargo to a safe hiding place. It likely plays a role in these beetles’ orientation to landmarks in the sky–which may be one way these beetles avoid rolling in circles and ending up right back where they started, in poop central. Signals like the rising or setting sun, the wind, and polarized moonlight are in the arsenal of dung beetles’ compass cues, without which they struggle to orient and travel in wobbly, circular paths.

And for at least one nocturnal species, the vital clue on a moonless night seems to be our own Milky Way. In 2013, a research team took some specimens of Scarabeus satyrus (and some poop) to a planetarium in Johannesberg, South Africa, and tested their navigational skills with and without a projection of our galaxy’s trademark diffuse band of light. Remarkably, the beetles steered straight–and floundered when fitted with tiny cardboard hats to block their eyes. 

Later research suggested these beetles are using the differences in brightness along the Milky Way, as opposed to individual star patterns, to find their way. And other research has found these nocturnal dung-haulers may be vulnerable to increasing light pollution, as bright lights provide confusing new landmarks, and diffuse light washes out the Milky Way’s vital contrasts.

Why does it matter if a beetle runs its poop in a straight line? Dung beetle success is crucial to healthily dispersing animal waste in ecosystems by bringing nutrients directly into the soil. There’s a citation-less figure floating around the internet that in some parts of Texas, dung beetles bury 80 percent of manure from cattle ranching–either way, these insects can absolutely demolish an individual cowpat.  And in so doing, they may help reduce methane and even CO2 emissions from livestock manure. 

And by burying and eating dung, they reduce the opportunities for flies and other pests to propagate–a lesson Australians learned the hard way when native dung beetles proved poorly adapted to assisting with the dung of colonists’ influx of cows and sheep. Thankfully for Australians, a government program to introduce and appreciate cow poop-eating beetles has been active since the 1960s.

FACT: Bananas might be the secret to better beer

By Laura Baisas 

Bananas might be the secret to a better beer. In 2022, a team of microbiologists in Belgium reported that they could improve contemporary beer’s flavor by genetically engineering a type of yeast. They focused on a gene for a banana-like flavor, “because it is one of the most important flavors present in beer, as well as in other alcoholic drinks.” The brewing of beer has a fascinating scientific, culinary, and sociological history, with women serving as brewmasters for centuries. Listen to this week’s episode to learn more! 

FACT: Made up animals can help us understand how language evolves online—and how resistant people are to censorship

By Rachel Feltman

In recent years, folks have started talking about a phenomenon often called “algospeak.” It’s where users of social media platforms like Instagram and TikTok come up with code words to help them avoid algorithms that suppress or outright ban certain topics of discussion. But this kind of linguistic innovation is far from new. In fact, researchers have been studying a similar tactic among Chinese internet users for more than a decade. 

The classic example of this phenomenon is discussion of an animal called the “grass mud horse.” Despite having millions of results on Google, the animal doesn’t exist. Its name is a sort of homophone you can only play with in a tonal language like Mandarin or Cantonese. By shifting tones, you turn cǎonímǎ—grass mud horse—into cào nǐ mā, which is a very profane insult. Talking about this made up animal (and an assortment of other imaginary species) allowed internet users to curse, discuss topics considered taboo by the algorithm, and criticize government officials and their policies without risking censorship. You can see a great visualization of how this kind of wordplay works here. 

In some cases this has led to seemingly random words being banned. For example, the name of the band Hoobastank became a stand-in for the censored word “tank,” and is now itself flagged! 

Chinese social media platform Weibo recently pledged to crack down on this kind of pun-based censorship evasion for good. But luckily they’ve got their work cut out for them, because people just keep upping their wordplay game. 

The post When steering balls of poop, dung beetles use the stars to navigate appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

Certain Airbnb hosts will need to make a few adjustments to their properties. On Monday, the short-term rental service announced it is finally prohibiting the use of all indoor security cameras, regardless of room location. For years, hosts could install video cameras in “common areas” such as living rooms, kitchens, and hallways, so long as they were both clearly visible and disclosed in the listings. Beginning April 30, however, zero such devices are permitted within any Airbnb location around the world.

Airbnb’s head of community policy and partnerships announced that the policy shift is intended to offer “new, clear rules” for both hosts and guests while providing “greater clarity about what to expect on Airbnb.” Privacy advocates have previously voiced concerns about what footage could be captured even in Airbnb “common spaces,” and are celebrating the news.

“No one should have to worry about being recorded in a rental,” Albert Fox Cahn, executive director of the civil rights watchdog nonprofit, Surveillance Technology Oversight Project (STOP), said in a public statement. STOP has campaigned Airbnb for this specific policy change since 2022. Cahn also called the policy reversal “a clear win for privacy and safety,” citing the allegedly easy exploitation of recording devices.

[Related: How to rent out your spare room and be an excellent host.]

According to the company, most Airbnb locales do not report indoor security cameras, so the upcoming policy revision is likely to only impact a smaller portion of rentals. And while indoor video cameras are soon-to-be banned, Airbnb will continue allowing other monitoring devices in rental locations under certain circumstances. Both doorbell and outdoor cameras, for example, are still permitted, so long as these are disclosed to guests and are not angled to see inside a residence. Cameras are also still prohibited from outdoor spots with “a greater expectation of privacy,” such as saunas or pool showers.

Other devices that remain available to hosts are decibel monitors to measure a common space’s noise levels—an increasingly popular tool meant to dissuade unauthorized parties. That said, the equipment must only be designed to assess sound volume, and can’t actually record or transmit audio.

After April 30, guests can report any hosts that do not adhere to the new regulations, with penalties including listing or account bans as a result.

The post Airbnb finally bans all indoor security cameras appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

This article was originally featured on Hakai Magazine, an online publication about science and society in coastal ecosystems. Read more stories like this at hakaimagazine.com.

In the summer of 2020, Jennifer Pytka spent three and a half hours a day sleuthing the internet for evidence of wildlife trafficking. She’d type กระเบนท้องน้ำ, a Thai word that loosely translates to stingray, into Google, and her search would immediately yield images of rings, each studded with an ornate white thorn about the size of a thumbnail. Pytka, a doctoral candidate at the Università di Padova in Italy, is investigating the previously undocumented trade of bowmouth guitarfish—a critically endangered ray whose spine and brows are adorned with these thorns. In Thailand, the horns are made into amulets, such as rings and bracelets, believed to have protective properties. In a 2023 study, Pytka notes how she pinpointed 977 of these items on online vending platforms, such as Facebook Marketplace, eBay, and the Alibaba-owned e-commerce site Lazada, over 21 days.

Bowmouth guitarfish amulets are just one example of the boundless number of protected wildlife products sold online, where a global Grand Bazaar of seedy vendors hawk their wildlife wares, and anyone with internet access can find products from rhino horns to exotic orchids to tiger claws with just a few clicks. With lax regulations, even weaker enforcement, and a lack of legal culpability, not only is wildlife trafficking able to fester online, but algorithms actually amplify sales, boosting the platforms’ profits.

Products sourced from protected species can be found across all manner of vending platforms, but with three billion active monthly users, Facebook is the grand pooh-bah. Pytka found 30 percent of the bowmouth guitarfish products on Facebook and 65 percent spread across other e‑commerce sites, such as Shopee and Lazada. “I’ve come to believe that Facebook is a driver of the global extinction crisis,” says Gretchen Peters, director of the Alliance to Counter Crime Online (ACCO), a nonprofit whistle-blower organization.

Prior to the emergence of the internet and online trading, vendors selling wildlife products had to connect with their customers largely through in-person networking, says David Roberts, a conservation scientist at the University of Kent in England who researches wildlife trafficking. But in the early 2000s, an increasing number of transactions in the physical world went digital, with wildlife trafficking being no exception. Today, nearly 6,000 species of plants and animals are traded illicitly, and the trafficking is worth up to US $23-billion annually. It is the fourth-largest illegal market, and many animals, such as rhinos, pangolins, and some species of parrots and sharks, are at risk of extinction due to their popularity on the black market.

The Convention on International Trade in Endangered Species of Wild Fauna and Flora (CITES) identifies at-risk species and designates protections and trade prohibitions. On-the-ground enforcement of CITES rules, however, is another matter.

Glenn Sant, a senior adviser on fisheries trade for TRAFFIC, a nonprofit aiming to reduce illegal trafficking, describes a hypothetical example of what might play out when someone catches a protected species of shark. “The fins will potentially be going to Hong Kong or China, and the meat might be going to Europe,” he says, adding that the skin might become leather and the oils sold for cosmetic products. Sant says that processing, shipping, and distribution around the world can make illicit animal harvesting nearly impossible to trace and therefore convict. That’s part of the reason Pytka chose to study bowmouth guitarfish—their unique thorns are easy to distinguish.

eBay was the first to acknowledge the growing problem of online trafficking and banned all ivory sales on its platform in 2009. Another milestone was reached in 2018 with the creation of the Coalition to End Wildlife Trafficking Online. This alliance, spearheaded by animal welfare groups TRAFFIC, the World Wildlife Fund (WWF), and the International Fund for Animal Welfare, advises technology platforms on how to identify and prevent wildlife trafficking. So far, 47 companies have joined the coalition, including Meta—the parent company of Facebook and Instagram—eBay, TikTok, and other international giants like Alibaba. The coalition’s most recent report, from 2021, found that between all the platforms, more than 11.6 million products made from prohibited wildlife have been removed or banned. A spokesperson from eBay said that over 350,000 listings for prohibited wildlife items were blocked or removed in 2022. Giavanna Grein, a wildlife specialist at WWF, encourages platforms to be more transparent with the public and concedes that the efforts undertaken by the coalition are just one small part of the picture. “We fully acknowledge this is a very complex and challenging issue, and there’s no one organization or effort that can tackle this,” she says.

Even with all the efforts, loopholes remain. Despite eBay’s ivory ban, for instance, a quick search by Roberts identified what he believes to be elephant ivory being sold under a code name. The product is still so readily available, in fact, that he centers his students’ projects on it. Similarly, a quick search on Facebook Marketplace for rhino horns for sale in southeast Asia immediately yields several posts.

Meta’s own policy prohibits “attempts to buy, sell, trade, donate, gift, or solicit endangered species or their parts,” and in a statement, a spokesperson said that content that violates their policies is removed. However, whistle-blower reports published since Facebook joined the Coalition to End Wildlife Trafficking Online have been scathing. “Facebook policy and public comments about countering illicit content are rendered virtually meaningless by the firm’s ineffective follow-up and enforcement,” reads a 2020 report from the ACCO. To assess the severity of wildlife trafficking on Facebook, the report used search terms such as “exotic + animal + for sale” in English, Arabic, Vietnamese, and Indonesian, turning up 473 Facebook pages and 281 groups openly selling wildlife products. Over half the pages were created since Facebook joined the coalition, showing that online trafficking appears to have increased.

In part, researchers were able to find so many illicit items because the Facebook algorithm is designed to recommend similar products and thus amplify the connections between vendors and prospective clients. (While looking at bowmouth guitarfish rings on Facebook Marketplace in Thailand, for instance, I saw posts for tiger claw amulets. After clicking to view them, my marketplace page automatically filled with curios made from guitarfish, tiger claw, and elephant ivory.) The ACCO report found 29 percent of the wildlife trafficking pages through Facebook’s “Related Pages” feature. Avaaz, a nonprofit that supports global activism, carried out a similar investigation and found that Facebook’s algorithm directed the researchers to dozens of wildlife groups, more than half of which contained potentially harmful wildlife trafficking content. Since it appears that Facebook’s algorithms are able to identify wildlife products, the algorithms should be able to hide these posts rather than promote them. When I asked about the discrepancy, Meta did not respond to this or any other question.

Peters says Meta is also passively profiting from the illegal activity. The platform makes money from embedded advertisements, and the online storefront Facebook Shops takes a small transaction fee from sales—including those of trafficked animals.

“[Facebook’s] platform is so big … and it’s in so many different languages that it’s really going to take a Herculean effort and a huge investment,” says Peters. “I don’t think Facebook is prepared to make the investments to clean up their own mess.” Peters also notes that Facebook could be more proactive in collaborating with law enforcement to dismantle criminal networks. “Facebook is sitting on a huge amount of information about some of the world’s biggest wildlife trafficking networks,” she says, and in many circumstances, the platform is not proactively showing that intelligence to law enforcement, claiming they’re protecting user privacy. Yet she says the firm is renowned for harvesting user data to sell to private companies. “It’s completely contradictory to me.” eBay is attempting to tackle this problem by implementing a regulatory portal that allows law enforcement authorities easy access to suspected criminal activity.

For the benefit of regular citizens looking to report posts on these platforms, I ask Roberts if taking down posts is akin to a game of whack-a-mole—with new posts cropping up as others are removed. “I don’t think we actually have the mallet to hit the mole,” replies Roberts.

In spite of the efforts of animal welfare and social justice groups like WWF and the ACCO, illicit wildlife sales are able to thrive online because platforms are protected from civil liability by section 230 of the Communications Decency Act in the United States. The act generally protects the platforms from being liable for the nefarious content they host.

“The way section 230 works is [that] any content created by a user like you or me or anybody else is considered free expression,” explains Peters. But she argues that illegal sales occurring over online platforms aren’t free speech—they’re felonies, and implementing something like a duty-of-care law would require platforms to remove criminal activity.

“I think [the platforms] should be held accountable,” says Roberts, who compares online trafficking to a bar allowing drugs to be sold in the bathrooms. The establishment is liable for allowing illicit activity on its premises. “How is that any different [from] a platform allowing illegal trade to take place?”

Both ACCO and Avaaz suggest simple measures for Facebook to reduce online wildlife crime. For example, when a user searches “bowmouth guitarfish amulets,” the algorithm could fail to return a search or trigger a pop-up explaining that the amulets come from a protected species. AI algorithms could also automatically flag questionable content or be used to trace trafficking activity. Pytka says it would be relatively simple to design such a system for bowmouth guitarfish rings because they’re so visually distinct. In early 2023, eBay acquired an AI-based software that will supposedly make the marketplace safer. In the meantime, though, my Facebook Marketplace home page swims with skeletal amulets, while researchers like Pytka can only speculate about how many of the endangered fish remain in the sea.

This article first appeared in Hakai Magazine and is republished here with permission.

The post How social media helps wildlife trafficking thrive in plain sight appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

There’s no doubt an extremely bright fireball careened through the atmosphere north of Papua New Guinea on January 8, 2014. It’s also true that divers recovered materials at the bottom of the ocean last year near where many experts believed the object landed—and that prominent Harvard astrophysicist Avi Loeb theorized some of these metallic spherules were possibly of “extraterrestrial technological” origin. But as to the ground vibrations recorded at a seismic station on Manus Island during the same atmospheric event? The explanation is likely much more mundane.

“[T]hey have all the characteristics we’d expect from a truck and none of the characteristics we’d expect from a meteor,” Johns Hopkins planetary seismologist Benjamin Fernando said on Thursday.

Fernando and his colleagues will present their findings on March 12 during the annual Lunar and Planetary Science Conference in Houston, Texas.

Although Fernando’s team concedes it’s difficult to prove what something isn’t through signal data, it’s pretty easy to highlight the characteristics it may share with existing, explainable seismic info. 

“The signal changed directions over time, exactly matching a road that runs past the seismometer,” said Fernando.

[Related: How scientists decide if they’ve actually found signals of alien life.]

To further bolster the much more everyday explanation, researchers also utilized data collected during the 2014 event by facilities in Australia and Palau originally built to measure nuclear test sound waves. After factoring in those recordings, Fernando’s team revised the previous location estimations for a more exact spot of the atmospheric occurrence—an area 100 miles away from the original region.

“The fireball location was actually very far away from where the oceanographic expedition went to retrieve these meteor fragments,” Fernando said of the 2023 recovery trip. “Not only did they use the wrong signal, they were looking in the wrong place.”

The team also doesn’t mince words in their new paper, “Probably Not Aliens: Seismic Data Analysis from the 2014 ‘Interstellar Meteor.’” Of the alien theory, the researchers “consider it to be at best highly overstated and at worst entirely erroneous.” And of the material recovered last year, “poor localisation implies that any material recovered is far less likely to be from the meteor, let alone of interstellar or even extraterrestrial origin.”

[Related: How lightning on exoplanets could make it harder to find alien life.]

Given NASA’s estimate that around 50 tons of meteoritic material bombards Earth every day, Fernando’s team says it’s definitely possible some of those fragments retrieved from the ocean floor may indeed be from some other meteorite. Regardless, they “strongly suspect that it wasn’t aliens.”

Disappointing? Perhaps. But there’ll probably be plenty of new UAP sightings to parse in the future—especially if people take up the government’s offer to submit their own inexplicable events.

For more detailed debunking, tune into a livestream of next week’s findings here.

The post ‘Alien’ signal was likely a very big truck appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

The Transportation Security Administration is launching the pilot phase of an autonomous self-screening checkpoint system. Unveiled earlier this week and scheduled to officially open on March 11 at Harry Reid International Airport in Las Vegas, the station resembles grocery store self-checkout kiosks—but instead of scanning milk and eggs, you’re expected to…scan yourself to ensure you aren’t a threat. Or at least that’s what it seems from the looks of it.

“We are constantly looking at innovative ways to enhance the passenger experience, while also improving security,” TSA Administrator David Pekoske said on Wednesday, claiming “trusted travelers” will be able to complete screenings “at their own pace.”

For now, the prototype station is only available to TSA PreCheck travelers. Although it’s possible additional passengers could use similar self-scan options in the future, depending on the prototype’s success. Upon reaching the Las Vegas airport’s “TSA Innovation Checkpoint,” users will see something similar to the standard security checks alongside the addition of a camera-enabled video screen. TSA agents are still nearby, but they won’t directly interact with passengers unless they request assistance, which may also take the form of a virtual agent popping up on the video screen.

The new self-guided station’s X-ray machines function similarly to standard checkpoints, while its automated conveyor belts feed all luggage into a more sensitive detection system. That latter tech, however, sounds a little overly cautious at the moment. In a recent CBS News video segment, items as small as a passenger’s hair clips triggered the alarm. That said, the station is designed to allow “self-resolution” in such situations to “reduce instances where a pat-down or secondary screening procedure would be necessary,” according to the TSA.

[Related: The post-9/11 flight security changes you don’t see.]

The TSA’s proposed solution to one of airports’ most notorious bottlenecks comes at a tricky moment for both the travel and automation industries. A string of recent, high-profile technological and manufacturing snafus have, at best, severely inconvenienced passengers and, at worst, absolutely terrified them. Meanwhile, businesses’ aggressive implementation of self-checkout systems has backfired in certain markets as consumers increasingly voice frustrations with the often finicky tech. Meanwhile, critics contend that automation “solutions” like the TSA’s new security checkpoint project are simply ways to employ fewer human workers who often ask for pesky things like living wages and health insurance.

Whether or not self-scanning checkpoints become an airport staple won’t be certain for a few years. The TSA cautioned as much in this week’s announcement, going so far as to say some of these technologies may simply find their way into existing security lines. Until then, the agency says its new prototype at least “gives us an opportunity to collect valuable user data and insights.”

And if there’s anything surveillance organizations love, it’s all that “valuable user data.”

The post TSA is testing a self-screening security checkpoint in Vegas appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

It took less than 30 minutes for Varda Space Industries’ W-1 capsule to leave its orbital home of eight months and plummet back to Earth. Such a short travel time not only required serious speed (around 25 times the speed of sound), but also the engineering wherewithal to endure “sustained plasma conditions” while careening through the atmosphere. In spite of these challenges, Varda’s first-of-its-kind reentry mission was a success, landing back on the ground on February 21. To celebrate, the company has released video footage of the capsule’s entire descent home.

Check out W-1’s fiery return below—available as both abbreviated and extended cuts:

Below is a longer 5-minute edit from separation to touchdown:

The full unedited raw footage and audio from separation to touchdown is available on our YouTube: https://t.co/ipdBvx93iB pic.twitter.com/ggIRHUvnnI

— Varda Space Industries (@VardaSpace) February 28, 2024

Installed on a Rocket Lab Photon satellite bus, Varda’s W-1 capsule launched aboard a SpaceX Falcon 9 rocket June 12, 2023. Once in low-Earth orbit, its mini-lab autonomously grew crystals of the common HIV treatment drug ritonavir. Manufacturing anything in space, let alone pharmaceuticals, may seem like overcomplicating things, but there’s actually a solid reason for it. As Varda explains on its website, processing materials in microgravity may benefit from a “lack of convection and sedimentation forces, as well as the ability to form more perfect structures due to the absence of gravitational stresses.”

In other words, medication crystals like those in ritonavir can be grown larger and more structurally sound than is typically possible here on Earth.

Although the experiment wrapped up in just three weeks, Varda needed to delay reentry plans multiple times due to issues securing FAA approval. After finally getting the go-ahead, the W-1 readied for its return earlier this month. All the while, it contained a video camera ready to capture its dramatic fall.

After ejecting from its satellite host, W-1 begins a slightly dizzying spin that provides some incredible shots from hundreds of miles above Earth. At about the 12-minute mark, the planet’s gravitational pull really takes hold—that’s when things begin to heat up for Varda’s experimental capsule.

[Related: First remote, zero-gravity surgery performed on the ISS from Earth (on rubber)]

At Mach 25 (around 17,500 mph), exterior friction between the craft and Earth’s atmosphere becomes so intense that it literally splits the chemical bonds of nearby air molecules. This results in a dazzling show of sparks and plasma before W-1’s parachute deploys to slow and stabilize its final descent. Finally, the capsule can be seen touching down in a remote region of Utah, where it was recovered by the Varda crew.

Next up will be an assessment of the space-grown drug ingredients, and additional launches of capsules for more manufacturing experiments. While they might not all include onboard cameras to document their returns, W-1’s is plenty mesmerizing enough.

The post Watch the plasma fly in space capsule’s dramatic fall to Earth appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

You probably don’t know about Automattic, but they know you.

As the parent company of WordPress, its content management systems host around 43 percent of the internet’s 10 million most popular websites. Meanwhile, it also owns a vast suite of mega-platforms including Tumblr, where a massive amount of embarrassing personal posts live. All this is to say that, through all those countless Terms & Conditions and third-party consent forms, Automattic potentially has access to a huge chunk of the internet’s content and data.

[Related: OpenAI’s Sora pushes us one mammoth step closer towards the AI abyss.]

According to 404 Media earlier this week, Automattic is finalizing deals with OpenAI and Midjourney to provide a ton of that information for their ongoing artificial intelligence training pursuits. Most people see the results in chatbots, since tech companies need the text within millions of websites to train large language model conversational abilities. But this can also take the form of training facial recognition algorithms using your selfies, or improving image and video generation capabilities by analyzing original artwork you uploaded online. It’s hard to know exactly what and how much data is used, however, since companies like Midjourney and OpenAI maintain black box tech products—such is the case in this imminent business deal.

So, what if you wanna opt-out of ChatGPT devouring your confessional microblog entries or daily workflows? Good luck with that.

When asked to comment, a spokesperson for Automattic directed PopSci to its “Protecting User Choice” page, published Tuesday afternoon after 404 Media’s report. The page attempts to offer you a number of assurances. There’s now a privacy setting to “discourage” search engine indexing sites on WordPress.com and Tumblr, and Automattic promises to “share only public content” hosted on those platforms. Additional opt-out settings will also “discourage” AI companies from trawling data, and Automattic plans to regularly update its partners on which users “newly opt out,” so that their content can be removed from future training and past source sets.

There is, however, one little caveat to all this:

“Currently, no law exists that requires crawlers to follow these preferences,” says Automattic.

The wild west of copyright and privacy

“From what I have seen, I’m not exactly sure what could be shared with AI,” says Erin Coyle, an associate professor of media and communication at Temple University. “We do have a confusing landscape right now, in terms of what data privacy rights people have.”

To Coyle, nebulous access to copious amounts of online user information “absolutely speaks” to an absence of cohesive privacy legislation in the US. One of the biggest challenges impeding progress is the fact that laws, by and large, are reactive instead of preventative regulation.

“It’s really hard for legislators to get ahead of the developments, especially in technology,” she adds. “While there are arguments to be made for them to be really careful and cautious… it’s also very challenging in times like this, when the technology is developing so rapidly.”

As companies like OpenAI, Google, and Meta continue their AI arms race, it’s the everyday people providing the bulk of the internet’s content—both public and private—who are caught in the middle. Clicking “Yes” to the manifesto-length terms and conditions prefacing almost every app, site, or social media platform is often the only way to access those services.

“Everything is about terms of service, no matter what website we’re talking about,” says Christopher Terry, a University of Minnesota journalism professor focused on regulatory and legal analysis of media ownership, internet policy, and political advertising.

Speaking to PopSci, Terry explains that basically every single terms of service agreement you have signed online is a legal contractual obligation with whoever is running a website. Delve deep enough into the legalese, and “you’re gonna see you agreed to give them, and allow them to use, the data that you generate… you allowed them to monetize that.”

Of course, when was the last time you actually read any of those annoying pop-ups?

“There is no data privacy in general,” Terry says. “With the digital lives that we have been living for decades, people have been sharing so much information… without really knowing what happens to that information,” Coyle continues. “A lot of us signed those agreements without any idea of where AI would be today.”

And all it takes to sign away your data for potential AI training is a simple Terms of Service update notification—another pop-up that, most likely, you didn’t read before clicking “Agree.”

You either opt out, or you’re in

Should Automattic complete its deal with OpenAI, Midjourney, or any other AI company, some of those very same update alerts will likely pop-up across millions of email inboxes and websites—and most people will reflexively shoo them away. But according to some researchers, even offering voluntary opt-outs in such situations isn’t enough.

“It is highly probable that the majority of users will have no idea that this is an option and/or that the partnership with OpenAI/Midjourney is happening,” Alexis Shore, a Boston University researcher focused on technology policy and communication studies, writes to PopSci. “In that sense, giving users this opt-out option, when the default settings allow for AI crawling, is rather pointless.”

Experts like Shore and Coyle think one potential solution is a reversal in approach—changing voluntary opt-outs to opt-ins, as is increasingly the case for internet users in the EU thanks to its General Data Protection Regulation (GDPR). Unfortunately, US lawmakers have yet to make much progress on anything approaching that level of oversight.

The next option, should you have enough evidence to make your case, is legal action. And while copyright infringement lawsuits continue to mount against companies like OpenAI, it will be years before their legal precedents are established. By then, it’s anyone’s guess what the AI industry will have done to the digital landscape, and your privacy. Terry compares the moment to a 19th-century gold rush.

“They’re going all in on it right now while they still can,” he says. “You’re going out there to stake out your claim right now, and you’re pouring everything you can into that machine so that later, when that’s a [legal] problem, it’s already done.”

 Neither OpenAI nor Midjourney responded to multiple requests for comment at the time of writing.

The post OpenAI wants to devour a huge chunk of the internet. Who’s going to stop them? appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

It turns out that last month’s report on Apple kicking its tortured, multibillion dollar electric vehicle project down the road another few years was a bit conservative. During an internal meeting on Tuesday, company representatives informed employees that all EV plans are officially scrapped. After at least a decade of rumors, research, and arguably unrealistic goals, it would seem that CarPlay is about as much as you’re gonna get from Apple while on the roads. RIP, “iCar.”

The major strategic decision, first reported by Bloomberg, also appears to reaffirm Apple’s continuing shift towards artificial intelligence. Close to 2,000 Special Projects Group employees worked on car initiatives, many of whom will now be folded into various generative AI divisions. The hundreds of vehicle designers and hardware engineers formerly focused on the Apple car can apply to other positions, although yesterday’s report makes clear that layoffs are imminent.

[Related: Don’t worry, that Tesla driver only wore the Apple Vision Pro for ’30-40 seconds’]

Previously referred to as Project Titan or T172, Apple’s intentions to break into the automotive market date as far back as at least 2014. It was clear from the start that Apple executives such as CEO Tim Cook wanted an industry-changing product akin to the iPod or iPhone—an electric vehicle with fully autonomous driving capabilities, voice-guided navigation software, no steering wheel or even pedals, and a “limousine-like interior.”

As time progressed, however, it became clear—both internally and vicariously through competitors like Tesla—that such goals were lofty, to say the least. Throughout multiple leadership shakeups, reorganizations, and reality checks, an Apple car began to sound much more like existing EVs already on the road. Basic driver components returned to the design, and AI navigation plans downgraded from fully autonomous to current technology such as acceleration assist, brake controls, and adaptive steering. Even then, recent rumors pointed towards the finalized car still costing as much as $100,000, which reportedly concerned company leaders for the hyper-luxury price point.

This isn’t the first time Apple pulled the plug on a major project—in 2014, for example, saw the abandonment of a 4K Apple smart TV. But the company has rarely, if ever, spent as much time and money on a product that never even officially debuted, much less made it to market.

Fare thee well, Apple Car. You sounded pretty cool, but it’s clear Tim Cook believes its future profits reside in $3,500 “spatial computing” headsets and attempting to integrate generative AI into everything. For now, the closest anyone will get to an iCar is wearing Apple Vision Pro while seated in a Tesla… something literally no one recommends.

The post The Apple Car is dead appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

Imagine a crime scene. Chances are, you’re also imagining someone dusting for fingerprints. Despite recent debates of whether fingerprint evidence is accurate and reliable, it can still prove extremely useful in certain situations, such as narrowing down potential suspect lists. Unfortunately, this technique often employs toxic powders, including environmentally harmful petrochemicals that can damage DNA evidence.

[Related: The racist history behind using biology in criminology.]

Thanks to a collaboration between scientists from the UK’s University of Bath and China’s Shanghai Normal University, this may change in the future. In a new study published in the Journal of the American Chemical Society, researchers laid out their case for a novel method of lifting latent fingerprints—a water soluble spray that is not only safer and faster, but easier to examine thanks to its ability to glow in the dark.

It all started with a tip-off from jellyfish.

For millions of years, many of these ocean invertebrates have contained Green Fluorescent Protein (GFP), which are fluorescent under certain lighting conditions. Knowing this, the team created two different dyes, LFP-Yellow and LFP-Red, that are based on the protein found in jellyfish. Short for “latent fingerprints,” LFP-Yellow and LFP-Red are applied using a simple spray bottle, which then selectively binds to negatively-charged molecules within fingerprints. Once stuck to the residual prints, the dyes begin to glow under blue light in just 10 seconds.

Interestingly, the solution is only “weakly fluorescent” before applied to LFPs, according to University of Bath researcher, Luling Wu, in a recent profile. It’s only once the dyes interact with fingerprint’s fatty or amino acids created by skin oil and sweat that they glow brighter.

Because it is applied as a fine mist, forensics examiners don’t need to worry about splashes that could potentially disturb prints. It also avoids the mess that often accompanies dusting with frequently toxic powders, and is even effective on rougher surfaces like concrete or brick.

Going forward, researchers hope to make their less harmful solution available commercially, as well as expand on the number of fluorescent colors to ensure use across a wider array of surfaces. Forensic analysts may not consider fingerprint evidence as ironclad as before, but with alternative methods of detection, they could soon lift them more accurately and safely. What’s more, doing so won’t risk damaging any nearby, much more sought after DNA clues.

The post Jellyfish-inspired glowing dye can glom onto fingerprints at crime scenes appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

Less than 4,500 tigers remain in the world, according to the International Union for the Conservation of Nature (IUCN). Habitat loss continues to pose an immense existential threat to the planet’s largest cat species—a problem compounded due to the animals residing in some of Earth’s most ecologically at-risk regions and landscapes.

To better monitor the situation in real time, NASA, Google Earth Engine, and over 30 researcher collaborators are announcing TCL 3.0 today, a new program that combines satellite imagery and powerful computer processing to keep an eye on tigers’ existing and reemerging ecosystems.

“The ultimate goal is to monitor changes in real time to help stabilize tiger populations across the range,” Eric W. Sanderson, VP for Urban Conservation at the New York Botanical Garden and first author of a recent foundational study published in Frontiers in Conservation Science explained.

[Related: A new algorithm could help detect landslides in minutes.]

“Tiger Conservation Landscapes,” or TCLs, refer to the planet’s distinct locales where Panthera tigris still roam in the wild. Because of their size, diet, and social habits, tigers require comparatively large areas to not only survive, but flourish.

According to researchers, stable tiger populations “are more likely to retain higher levels of biodiversity, sequester more carbon, and mitigate the impacts of climate change, at the same time providing ecosystem services to millions of humans in surrounding areas.” In doing so, TCLs can serve as a reliable, informative indicator of overall environmental health markers.

Unfortunately, the total area of Tiger Conservation Landscapes declined around 11 percent between 2001 and 2020. Meanwhile, potential restored habitats have only plateaued near 16 percent of their original scope—if such spaces were properly monitored and protected, however, tigers could see a 50 percent increase in available living space. 

Using this new analytical computing system based on Google Earth Engine data, NASA Earth satellite observations, biological info, and conservation modeling, TCL 3.0 will offer environmentalist groups and national leaders critical, near-real time tools for tiger conservation efforts.

“Analysis of ecological data often relies on models that can be difficult and slow to implement, leading to gaps in time between data collection and actionable science,” Charles Tackulic, a research statistician with the US Geological Survey, said in today’s announcement. “The beauty of this project is that we were able to minimize the time required for analysis while also creating a reproducible and transferable approach.”

Researchers say government and watchdog users of TCL 3.0 will be able to pinpoint tiger habitat loss as it happens, and hopefully respond accordingly. National summaries of initial available data can be found through the Wildlife Conservation Society, with more information to come.

TCL 3.0 provides an unprecedentedly complex and advanced monitoring system for one of the planet’s most threatened creatures, but as researchers note in their new study, the solution is arguably extremely simple.
“What have we learned about tiger conservation over the last two decades? Conservation works when we choose to make it so,” the authors conclude in their recent report. “Conservation is straightforward. Don’t cut down their habitat. Don’t stalk them, harass them, or kill them or their prey. Control poaching and extinguish the illegal trade in tiger bones and parts. Prevent conflicts with people and livestock wherever possible, and where and when not, then mitigate losses to forestall retaliation.”

Correction 2/27/24 5:53PM: This article has been updated to more accurately reflect the world’s remaining tiger population. PopSci regrets the error.

The post NASA and Google Earth Engine team up with researchers to help save tigers appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

Japan Aerospace Exploration Agency (JAXA) announced on Monday that its historic Smart Lander for Investigating Moon has defied the odds—after surviving a brutal, two-week lunar night while upside down, SLIM’s solar cells subsequently gathered enough energy to restart the spacecraft over the weekend. In an early morning post to X, JAXA reported it briefly established a communication relay with its lunar lander on Sunday, but the moon’s extremely high surface temperature currently prevents engineers from doing much else at the moment. Once SLIM’s instrument temperatures cool off in a few days’ time, however, JAXA intends to “resume operations” through additional scientific observations as long as possible.

[Related: This may be SLIM’s farewell transmission from the moon.]

SLIM arrived near the moon’s Shioli crater on January 19, making Japan the fifth nation to ever reach the lunar surface. Although JAXA’s lander successfully pulled off an extremely precise touchdown, it did so upside down after its main engines malfunctioned about 162-feet above the ground. The resulting nose-down angle meant SLIM’s solar cell arrays now face westward, thereby severely hindering its ability to gather power. Despite these problems, the craft’s two tiny robots still deployed and carried out their reconnaissance duties as hoped and snapped some images of the inverted lander. Meanwhile, SLIM transmitted its own geological survey data back to Earth for a few precious hours before shutting down.

Although JAXA officials cautioned that might be it for their lander, SLIM defied the odds and rebooted 10 days later with enough juice to continue surveying its lunar surroundings, such as identifying and measuring nearby rock formations.

“Based on the large amount of data obtained, analysis is now underway to identify rocks and estimate the chemical composition of minerals, which will help to solve the mysteries surrounding the origin of the Moon. The scientific results will be announced as soon as they are obtained,” JAXA said at the time.

But by February 1, the moon’s roughly 14.5-day lunar night was setting in, plunging temperatures down to a potentially SLIM-killing -208 Fahrenheit. Once again, JAXA bid a preemptive farewell to their plucky, inverted technological achievement—only to be surprised yet again over the weekend.

In the few days since the most recent lunar evening’s conclusion, SLIM apparently recharged its solar cells enough to come back online. But as frigid as the moon’s night phases are, its daytime temperatures can be just as brutal. According to JAXA, some of the lander’s equipment initially warmed up to over 212-degrees Fahrenheit. To play it safe, mission control is giving things a little time to cool off before tasking SLIM with additional scans, such as using its Multi-Band Camera to assess nearby regolith formations’ chemical compositions.

JAXA has a few more days before the moon enters another two-week night, during which SLIM will go into yet another hibernation. While it could easily succumb to the lunar elements this next time, it’s already proven far more resilient than its designers thought possible. It may not surpass expectations as dramatically as NASA’s Mars Ingenuity rotocopter (RIP), but the fact that SLIM made it this long is cause enough for celebration.

The post SLIM lives! Japan’s upside-down lander is online after a brutal lunar night appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

US farmers are closer than ever to raising genetically edited pigs immune to one of the animal’s deadliest diseases. But while millions of dollars could be saved with livestock impervious to highly virulent, diverse strains of porcine reproductive and respiratory syndrome (PRRS), animal rights groups maintain the cutting-edge idea isn’t an ethical solution, but yet another industrial farming stopgap.

PRRS is a dynamic, often fatal virus that affects millions of pigs around the world and costs farmers as much as $2.7 billion each year. Current vaccines only reduce symptom severity, and the antibiotics used to treat an infected pig’s weakened immune system can exacerbate the development of other resistant bacterial diseases.

[Related: Scientists swear their lab-grown ‘beef rice’ tastes ‘pleasant’]

Genus, an international breeding company, believes the best way to solve this major issue is to engineer pigs that are incapable of contracting the virus. As highlighted in a recent New Scientist profile, Genus researchers succeeded through CRISPR gene editing technology. By removing a portion of protein called CD163, the disease cannot infect a pig’s cells and allows the animal to remain “healthy and indistinguishable in appearance and behavior,” according to a Genus research study recently published in The CRISPR Journal.

Doing so isn’t an easy task. Just one-fifth of Genus-bred piglets possessed the desired gene—and even then, only within certain body cells due to a biological condition known as mosaicism. Meanwhile, some lab livestock may have lacked CD163, but at the cost of other unwanted genome changes.

Because of such issues, experts have spent years attempting to create a healthy, gene-edited animal. Genus says it has so far bred hundreds of PRRS-immune pigs, and expects to receive approval from the US Food and Drug Administration to begin public sales as soon as next year. Meanwhile, regulatory approvals are also being pursued globally in countries including China and Mexico—both of which import large amounts of US pork.

But according to factory farming critics and animal rights advocates, the real issue isn’t livestock disease susceptibility—it’s the livestock’s living conditions. According to the international welfare nonprofit World Animal Protection, farm stock receives three-quarters of global antibiotic supplies each year in an attempt to stave off disease, treat infections, and promote faster growth rates. Doing so is directly linked to the rise in treatment-resistant superbugs, which are more likely to leap between animals to humans within a factory farm’s cramped, poorly ventilated environments.

“Crowding animals into stressful, unhealthy conditions has led to the emergence of new virulent pathogens and diseases,” Gene Baur, President and Co-Founder of the animal rights group Farm Sanctuary, said in an email to PopSci. “Rather than developing genetically engineered animals who can survive the horrific cruelties of factory farming, agribusiness should focus instead on addressing the conditions that create these diseases in the first place.”

The post Gene-edited pigs immune to deadly virus could arrive on farms by next year appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

Facing bias accusations, Google this week was forced to pause the image generation portion of Gemini, its generative AI model. The temporary suspension follows backlash from users who criticized it for allegedly placing too much emphasis on ethnic diversity, sometimes at the expense of accuracy. Prior to Google pausing services, Gemini was found producing racially diverse depictions of World War II-era Nazis, Viking warriors, the US Founding Fathers, and other historically white figures.

loving google’s ongoing commitment to diversity here pic.twitter.com/kexIJImB2D

— pagliacci the hated 🌝 (@Slatzism) February 21, 2024

I asked it to generate images of the King of England the other day 🤣 pic.twitter.com/NEOqDELwmz

— Alexander Leishman 🇺🇸 (@Leishman) February 20, 2024

In a statement released Wednesday, Google said Gemini’s image generation capabilities were “missing the mark” and said it was “working to improve these kinds of depictions immediately.” Google then suspended access to the image generation tools altogether on Thursday morning and said it would release a new version of the model soon. Gemini refused to generate any images when PopSci tested the service Thursday morning, instead stating: “We are working to improve Gemini’s ability to generate images of people. We expect this feature to return soon and will notify you in release updates when it does.” As of this writing, Gemini is still down. Google directed PopSci to its latest statement when reached for comment.

Gemini’s image generations draw complaints  

Google officially began rolling out its image generation tools in Gemini earlier this month but controversy over its non-white depictions heated up this week. Users on X, formerly Twitter, began sharing screenshots of examples where Gemini reportedly generated images of nonwhite people when specifically prompted to depict a white person. In other cases, Gemini reportedly appeared to over-represent non-white people when prompted to generate images of historical groups that were predominantly white, critics claim.

The posts quickly attracted the attention of right-wing social media circles which have taken issue with what they perceive as heavy-handed diversity and equity initiatives in American politics and business. In more extreme circles, some accounts used the AI-generated images to stir-up an unfounded conspiracy theory accusing Google of purposely trying to eliminate white people from Gemini image results. 

How has Google responded to the Gemini controversy? 

Though the controversy surrounding Gemini seems to stem from critics arguing Google doesn’t place enough emphasis on white individuals, experts studying AI have long said AI models do just the opposite and regularly underrepresented nonwhite groups. In a relatively short period of time, AI systems trained on culturally biased datasets have amassed a history of repeating and reinforcing stereotypes about racial minorities. Safety researchers say this is why tech companies building AI models need to responsibly filter and tune their products. Image generators, and AI models more broadly, often repeat or reinforce culturally biased data absorbed from its training data in a dynamic researchers sometimes refer to as “garbage in, garbage out.” 

A version of this played out following the rollout of OpenAI’s DALL-E image generator in 2022 where AI researchers criticized the company for allegedly reinforcing age-old gender and racial stereotypes. At the time, for example, users asking DALL-E to produce images of a “builder” or a “flight attendant” would produce results exclusively depicting men and women respectively. OpenAI has since made tweaks to its models to try and address these issues. 

[ Related: How this programmer and poet thinks we should tackle racially biased AI ]

It’s possible Google was attempting to counterbalance some biases with Gemini but made an overcorrection during that process. In a statement released Wednesday, Google said Gemini does generate a wide range of people, which it said is “generally a good thing because people around the world use it.” Google did not respond to PopSci’s requests for comment asking for information on why Gemini may have produced the image results in question.

It’s not immediately clear what caused Gemini to produce the images it did but some commentators have theories. In an interview with Platformer’s Casey Newton Thursday, former OpenAI head of trust and safety Dave Willner said balancing how AI models responsibly generate content is complex and Google’s approach “wasn’t exactly elegant.” Wilners suspected these missteps could be attributed, at least in part, to a lack of resources provided to Google engineers to approach the nuanced area properly.

[ Related: OpenAI’s Sora pushes us one mammoth step closer towards the AI abyss ]

Gemini Senior Director of Product Jack Krawczyk elaborated on that further in a post on X, where he said the model’s non-white depictions of people reflected the company’s “global user base.” Krawczyk defended Google’s approach towards representation and bias, which he said aligned with the company’s core AI principles, but said some “inaccuracies” may be occuring in regard to historical prompts.

“We are aware that Gemini is offering inaccuracies in some historical image generation depictions, and we are working to fix this immediately,” Krawczyk said on a now restricted account. “Historical contexts have more nuance to them and we will further tune to accommodate that,” he added. 

Neither Krawczyk nor Google immediately responded to criticisms from users who said the racial representation choices extended beyond strictly historical figures. These claims, it’s worth stating, should be taken with a degree of skepticism. Some users expressed different experiences and PopSci was unable to replicate the findings. In some cases, other journalists claimed Gemini had refused to generate images for prompts asking the AI to create images of either Black or white people. In other words, user experiences with Gemini in recent days appear to have varied widely. 

Google released a separate statement Thursday saying it was pausing Gemini’s image generation capabilities while it worked to address “inaccuracies” and apparent disproportionate representations. The company said it would re-release a new version of the model “soon” but didn’t provide any specific date. 

We're already working to address recent issues with Gemini's image generation feature. While we do this, we're going to pause the image generation of people and will re-release an improved version soon. https://t.co/SLxYPGoqOZ

— Google Communications (@Google_Comms) February 22, 2024

Though it’s still unclear what caused Gemini to generate the content that resulted in its temporary pause, the flavor of online blowback Google received likely won’t end for AI-makers anytime soon.

The post Google pauses Gemini image tools after AI generates ‘inaccuracies’ in race appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

Earlier this week, Delta Air Lines announced an extra flight for its April 8 schedule, timed specifically to provide passengers an aerial view of the total solar eclipse. But if you were still hoping to snag a ticket for the afternoon jaunt alongside the path of totality, you’re already out of luck—seats aboard the Airbus A220-300 sold out within 24 hours.

According to Delta’s original announcement, DL Flight 1218 with service from Austin to Detroit will depart at 12:15 PM CST for its roughly 1,380-mile, 3-hour-long trip. Once at a cruising altitude of 30,000-feet, passengers will be able to view the celestial event through the plane’s “extra-large windows,” which the official Airbus specs manual says measure in at 11×16 inches. For comparison, a Boeing 777 includes 10×15 inch glimpses of the outside world. Everyone on the plane will receive special glasses to safely watch the eclipse (which is nice to hear, given how few free amenities remain on most commercial flights).

[Related: We can predict solar eclipses to the second. Here’s how.]

While the solar eclipse will last several minutes for anyone on the ground, Flight 1218’s timing and route should grant a longer spectacle.

As cool as a first-class seat to the eclipse would be, there are plenty of (likely cheaper) locations across the US to consider visiting on April 8. After traveling across Central America, the path of totality will pass across large portions of  Oklahoma, Arkansas, Missouri, Illinois, Kentucky, Indiana, Ohio, Pennsylvania, New York, Vermont, New Hampshire, and Maine.

If you’re truly determined to head to skies, NPR notes that there are other flight options scheduled to pass by at least some part of the eclipse, including from Delta, as well as several from Southwest.

But keep in mind: A plane’s altitude doesn’t necessarily guarantee a picture-perfect view of the eclipse—if anything, there’s a chance that cloud coverage could impede an onlooker’s vantage. There’s also the possibility of weather or air traffic control delays, which… well, this country has a history of such headaches.

So despite the multiple jetset options, your best bet to see April’s eclipse is simply making sure you’re within its route, firmly on the ground, and equipped with proper eyewear. Seriously, take it from NASA: “Viewing any part of the bright Sun through a camera lens, binoculars, or a telescope without a special-purpose solar filter secured over the front of the optics will instantly cause severe eye injury.”

The post Delta’s solar eclipse flight sold out, but your best bet to see it is still down here appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

Even in a digital-first world, optical disks like DVDs and Blu-rays still have their many uses. But despite being cheap, sturdy, and small, they can’t keep up with today’s storage needs. This is because, spatially speaking, optical disks almost always offer just a single, 2D layer–that reflective, silver underside–for data encoding. If you could boost a disk’s number of available, encodable layers, however, you could hypothetically gain a massive amount of extra space.

Researchers at the University of Shanghai for Science and Technology recently set out to do just that, and published the results earlier this week in the journal, Nature. Using a 54-nanometer laser, the team managed to record a 100 layers of data onto an optical disk, with each tier separated by just 1 micrometer. The final result is an optical disk with a three-dimensional stack of data layers capable of holding a whopping 1 petabit (Pb) of information—that’s equivalent to 125,000 gigabytes of data.

[Related: Inside the search for the best way to save humanity’s data.]

This is a bonkers amount of data compared to what can currently reside on even the most high-end flash or hybrid hard drives (HHDs). As Gizmodo offers for reference, that same petabit of information would require roughly a six-and-a-half foot tall stack of HHD drives—if you tried to encode the same amount of data onto Blu-rays, you’d need around 10,000 blank ones to complete your (extremely inefficient) challenge.

To pull off their accomplishment, engineers needed to create an entirely new material for their optical disk’s film, known as (take a big breath here) “dye-doped photoresist with aggregation-induced emission luminogens.” For brevity’s sake, AIE-DDPR is apparently just fine, too. AIE-DDPR film utilizes a combination of specialized, photosensitive molecules capable of absorbing photonic data at a nanoscale level, which is then encoded using a high-tech dual-laser array.

Because AIE-DDPR is so incredibly transparent, designers could apply layer-upon-layer to an optical disk without worrying about degrading the overall data. This basically generated a 3D “box” for digitized information, thus exponentially raising the normal-sized disk’s capacity.

But how much is a petabit, really? According to ZME Science, datasets used to train generative AI can include roughly 5.8 billion indexed webpages, totalling about 56 Pb of data. So, hypothetically, instead of relying on unsustainably energy-hungry data centers, one could conceivably fit all of ChatGPT’s training material in one of those retro CD album trapper keepers from the 2000s.

Unfortunately, a CD folder containing enough data to train your own AI program isn’t likely to arrive anytime soon. Creating the cutting-edge optical disk reportedly takes quite a while, and is still comparatively energy inefficient. Still, researchers believe they can solve for both hindrances with further experimentation and innovation. If so, some of the biggest issues in modern data management could be tackled by literally building upon a decades’ old physical format.

The post This DVD-sized disk can store a massive 125,000 gigabytes of data appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

You know the nightmare situation: You dropped your iPhone in water—be it pool, ocean, or toilet. Although iPhone 12’s and onward are designed to survive 30 minutes of aquatic submersion as deep as 20 feet, your worries get the best of you. In a frantic bid to save your expensive device from potential damage or even demise, you remember your friend’s suggestion to throw it in a bag of rice overnight. Supposedly, the grain draws out any remaining water droplets from the smartphone’s tiny crevices, saving its precious circuitry in the process. They swore by it, after all. What is there to lose?

[Related: Apple’s newest gadgets include titanium iPhones with USB-C ports.]

Well, whatever the supposed results (and despite a fair amount of longstanding contradictory evidence) the DIY repair is officially obtaining “urban myth” status. As MacWorld spotted earlier today, a recently updated Apple support document states in no uncertain terms that the ol’ bag of rice trick is bogus. What’s more, it could actually cause further issues in your iPhone.

“Don’t put your iPhone in a bag of rice,” Apple warns in the revised article on its dreaded Liquid Detection Alert. “Doing so could allow small particles of rice to damage your iPhone,” while the rice starch can gunk up the innards after making its way through the device’s small crevices. Besides all that, rice simply isn’t as effective a desiccant as other materials, such as those silica packets you already should be recycling, anyway.

Among the other rumored solutions to avoid, the company advises iPhone owners not to use an “external heat source” such as a blow dryer, as well as leave the compressed air can in the utility closet. Similarly, trying to stuff cotton swabs, napkins, paper towels, or any other “foreign object” into charging ports could make things worse.

So, what should you do if your iPhone takes a plunge? Apple advises a gentle approach in such situations, such as simply tapping the device against your hand “with the connector facing down” to dislodge liquid, then leaving it in an open, dry space with decent airflow for at least 30 minutes. From there, try connecting it to a cable charger.

If the Liquid Detection Alert proves persistent, Apple suggests allowing up to 24 hours to fully dry. And if even that doesn’t work? Well…

“If your phone has dried out but still isn’t charging, unplug the cable from the adapter and unplug the adapter from the wall (if possible) and then connect them again,” says Apple.

Yes. They really did pull out the trusty “Have you tried turning it on and off again?” line for this one.

The post Do not put your wet iPhone in rice, warns Apple appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

A 5,000-pound dead satellite resembling a spaceship from Star Wars is hurtling towards Earth, but don’t worry—experts say situations like this happen “every week or two.”

Launched in 1995 by the European Space Agency from Kourou, French Guiana, the European Remote Sensing 2 (ERS-2) array spent over a decade-and-a-half observing the planet’s topography and weather events, including natural disasters in remote, hard-to-document regions. Alongside its sibling, ERS-1, the pair were considered the “most sophisticated Earth observation spacecraft” ever developed at the time of their deployment.

ERS-2 spotted! 📸🛰️

The ESA satellite is on a tumbling descent that will lead to its atmospheric reentry and break up this week.

These images of ERS-2 were captured by @heospace for @spacegovuk using cameras on board other satellites.#ERS2reentry pic.twitter.com/GTuubP6apJ

— ESA Operations (@esaoperations) February 19, 2024

In July 2011, however, the ESA decided to retire its “nominally” functioning ERS-2 and begin a scheduled deorbiting process. The satellite underwent 66 maneuvers over the ensuing month, using up its remaining fuel to descend from an altitude from roughly 487-miles to 356-miles above the Earth’s surface. Since then, ERS-2’s orbit has slowly decayed to its current point—caught in the planet’s gravitational pull, and picking up speed as it falls into the atmosphere. 

On Sunday, the ESA posted grainy, black-and-white images to X taken last month by the Australian commercial imaging company, HEO, which show ERS-2 (then about 150-miles high) spiraling downwards during its final journey. From the camera’s vantage, the satellite certainly looks a lot like an incoming TIE Fighter from Star Wars. 

But no need to evade Imperial scrutiny—or even fiery orbital debris, for that matter. ERS-2 is currently falling at a rate of over 6.2 miles per day, a speed expected to accelerate as atmospheric drag takes an even greater hold. As of February 20, ERS-2 has around 120-or-so miles left to go, and will start breaking up and bursting into flames once about 50 miles high. Most, if not all, of the subsequent detritus will then immolate to harmless dust and ash, posing an extremely low damage risk for anything or anyone below it.

[Related: Some space junk just got smacked by more space junk, complicating cleanup.]

The ESA estimates ERS-2 will burn away around 3:53PM EST on Wednesday, although trackers offer as much as a 7-hour window on either side to account for “unpredictable solar activity” that could influence its descent speed. As to where in the world the satellite will fall apart—well, that part is a little more difficult to predict at the moment, although more accurate geolocation estimates are expected over the next day.

Deorbiting satellites is vital to ensuring enough room is kept for the thousands upon thousands of other human-made objects orbiting Earth. Increasingly crowded skies is a major concern for space agencies, private companies, and watchdog groups—an issue that isn’t likely to diminish anytime soon. Back in October, for example, a space junk cleanup mission proved more complicated when another piece of debris smacked into the satellite targeted for decommissioning. In the meantime, regulators like the FCC are fining companies for failing to do their part in accounting for their dead satellites.

After all, while a single satellite burning up during deorbit isn’t cause for concern—a “Kessler cascade” most certainly is. 

The post Dead satellite hurtles towards Earth in new grainy images appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

This article was originally featured on Hakai Magazine, an online publication about science and society in coastal ecosystems. Read more stories like this at hakaimagazine.com.

Few social networking platforms are known for inspiring positive social change these days, but an Inuit-developed app is helping Indigenous communities from Alaska to Greenland advance their self-determination. Named SIKU after the Inuktitut word for “sea ice,” the app allows communities in the North to pull together traditional knowledge and scientific data to track changes in the environment, keep tabs on local wild foods, and make decisions about how to manage wildlife—all while controlling how the information is shared.

A group of Inuit elders and hunters from Sanikiluaq, Nunavut, came up with the idea for SIKU more than a decade ago to document and understand the changing sea ice they were witnessing in southeastern Hudson Bay. The group turned to the local nonprofit Arctic Eider Society to develop a web-based platform where hunters in nearby coastal communities could upload photos and videos and share knowledge. Contributors began using the portal in 2015 to log water temperature and salinity data, note observations of important wildlife species—such as beluga and common eider ducks—and track the flow of contaminants through the food web.

Over the years, SIKU has evolved, and recently, the elders saw that the platform could help address a familiar challenge: sharing knowledge with younger people who often have their noses in their phones. In 2019, SIKU relaunched as a full-fledged social network—a platform where members can post photos and notes about wildlife sightings, hunts, sea ice conditions, and more. The app operates in multiple languages, such as Inuktitut, Cree, Innu, and Greenlandic, and includes maps with traditional place names. Since early 2024, over 25,000 people from at least 120 communities have made more than 75,000 posts on SIKU.

Members’ photos demonstrate the breadth and bounty of northern foods: They show plump bags of berries sitting on the tundra, clusters of sea urchins nestled on smooth gray stones, and boxes of fresh Arctic char placed in the snow. They depict harp seals, ringed seals, ptarmigan, beluga, common eider, and neat rows of colorful eggs laid out next to smiling kids. The posts tell stories of hunting and traveling, the impacts of climate change and industrial activity, and the migrations, diets, and illnesses of local animals. In effect, SIKU captures everyday Indigenous life in a rapidly changing landscape.

Traditionally, Inuit communities shared this information orally. “We have lived in the environment for centuries. We know about the wildlife,” says Lucassie Arragutainaq, a manager at the Sanikiluaq Hunters and Trappers Association and cofounder of the Arctic Eider Society. Yet industry representatives and government scientists have a long history of dismissing Indigenous knowledge and making decisions based on sparse environmental data collected during irregular, short-term studies. Now armed with SIKU, northerners are documenting information “in a way that [other] people will understand,” Arragutainaq says.

The app is also equipped with useful tools for life on the ice, including weather reports, sea ice forecasts, and other critical safety information. Hunters and harvesters can use their phones’ GPS to track their routes and geolocate each post and photo. “When I go out on the land with family, we go a long distance, and the SIKU app can show which area we are in. It’s precise,” says Karen Nanook, who lives in Taloyoak, Nunavut.

In June 2023, for instance, Nanook was heading home across the frozen ocean after an ice fishing trip when a rift appeared to open in the ice beneath one of her sled’s runners. “I thought the sled was going to fall in,” she says. But clear ice was covering the crack, and the sled stayed upright. After her close call, Nanook snapped a photo, tagged it as a “dangerous ice observation,” and posted it to SIKU to warn others.

The data held in SIKU is robust and up to date, and communities are already using the app to inform important decisions. In 2021, for example, elders in Sanikiluaq were worried the local reindeer population had thinned, so the Hunters and Trappers Association used SIKU to survey hunters and look at recent reported harvest rates. The analysis led the association to temporarily close the hunt to relieve pressure on the population and to reintroduce hunting slowly once the number of reindeer increased. This decision shows how Inuit can use the technology in combination with traditional wildlife management, says Arragutainaq. Today, the community is also using SIKU data to guide the development of the Qikiqtait Protected Area around the Belcher Islands, where Sanikiluaq is located.

SIKU has become the main tool for other research projects, too. “Having the people who are already the eyes and ears of the land use the platform to share that information will revolutionize the way we make decisions,” says Stephanie Varty, a wildlife management biologist at the Eeyou Marine Region Wildlife Board in the traditional territory of the Crees of Eeyou Istchee, in James Bay, Quebec.

Varty says trappers and land users from Eeyou Istchee’s five coastal communities—Waskaganish, Eastmain, Wemindji, Chisasibi, and Whapmagoostui—will soon use SIKU to document climate change in their region. They’ll also log observations and hunting stories, which will help the communities assess the environmental impacts of future development projects, including a proposed deep-sea port that would allow mining companies to access lithium and other minerals in the region.

Northern Indigenous communities are showing southerners that traditional knowledge should be taken seriously. “When Inuit knowledge is mobilized into graphs and diagrams, that [information] can’t be ignored and written off as anecdotal stories,” says Joel Heath, the executive director and cofounder of the Arctic Eider Society.

The ingenuity of SIKU is how it weaves together all kinds of insights about life in the North and supports community-driven research. “It’s part science and part Inuit knowledge,” says Arragutainaq. “It can work both ways, instead of one dominating the other.”

This article first appeared in Hakai Magazine and is republished here with permission.

The post On this social network, sea ice, traditional foods, and wildlife are always trending appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

Remember the old reality show competition stunt of getting contestants to eat live bugs on primetime television? Consuming “food” while it’s still alive spans numerous cultures around the world. In Japan, for example, odorigui (or “dance-eating”) is a centuries’ old tradition often involving squid, octopus, and tiny translucent fish known as ice gobies. Diners pop these still-living creatures into their mouths, as the wriggling is part of the overall meal experience.   

To potentially better understand the psychology and emotional responses associated with consuming odorigui dishes, researchers designed their own stand-in—a moving gelatin robo-food combining 3D-printing, kitchen cooking, and air pumps. The results appear not only tastier than your average reality show shock snack, but a potential step towards creative culinary and medical applications.

… And yet, judging from this video, it’s undeniably still a little odd.

Detailed in a study published earlier this month in PLOS One, a team at Japan’s University of Electro-Communications and Osaka University recently devised a pneumatically-driven handheld device to investigate what they dub “human-edible robot interaction,” or HERI. For the “edible” portion of HERI, researchers cooked up a gummy candy-like mixture using a little extra sugar and apple juice for flavor. 

After letting the liquid cure in molds that included two hollow airways, the team then attached the snack to a coffee mug-like holder. The design allowed researchers to inject air through the gelatin in different combinations—alternating airflow between each tube produced a side-to-side wagging motion, while simultaneous inflation offered a (slightly unnerving) pulsating movement.

And then, the taste tests.

The team directed 16 Osaka University students to grab the device holding their designated, writhing soft robot morsel, place the edible portion in their mouth, allow it to move about for 10 seconds, then chomp. Another (possibly relieved) group of control students also ate a normal, immobile gelatin gummy. Following their meals, each volunteer answered a survey including questions such as:

– Did you think what you just ate had animateness?

– Did you feel an emotion in what you just ate?

– Did you think what you just ate had intelligence?

– Did you feel guilty about what you just ate?

Perhaps unsurprisingly, it seems that a meal’s experience can be influenced by whether or not the thing you just put in your mouth is also moving around in your mouth. Students described this sensation using the Japanese onomatopoeic terms gabu, or “grappling,” and kori-kori, meaning “crisp.” Movement also more frequently caused volunteers to feel a bit of guilt at eating a “still living” dish, as well as attach a sense of intelligence to it.

[Related: Scientists swear their lab-grown ‘beef rice’ tastes ‘pleasant’]

While only an early attempt at looking into some of the dynamics in odorigui, researchers believe more intricate soft robot designs can allow for more accurate experiments. Meanwhile, such research could lead to a “deepening understanding of ethical, social, and philosophical implications of eating,” as well as potential uses in medical studies involving oral and psychological connections. There’s also a possibility for “innovative culinary” experiences down the line, so who knows what might be coming to high-brow restaurants in the future—perhaps gyrating gyros, or wobbly waffles. Hopefully, nothing too macabre will wind up on menus. It’s certainly something researchers took into consideration during their tests.

“NOTE: During the experiment, we did not draw a face on the edible robot,” reads the fine print at the bottom of the demonstration video, presumably meaning they were just having a bit of fun with the project.

Which is good to hear. Otherwise, this whole thing might have come across as weird.

The post This edible, wriggling robot mimics experience of eating moving food appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

The whole point of lab-grown meat, by and large, is to create a sustainable product capable of… you know, replacing meat. Researchers at universities and startup companies across the world have spent years and a lot of money on attempts to accurately imitate chicken, beef, fish, and even extinct woolly mammoths.

It’s an uphill battle, but convincing a substantial portion of the population to reduce, if not entirely cut, animal meat from their diets is widely considered a key way to combat industrial farming’s massive global carbon emissions. But instead of trying to replicate the minutiae of a burger’s mouthfeel and flavor, one group of scientists decided to sidestep those goals entirely for a new dish: “beef rice” grown from lab-cultured cow fat cells.

But if you are skeptical at the thought of spoonfuls of synthetic meat-grain meals, fear not: Its makers swear their pinkish globules offer its consumers a “unique blend of aromas” including that “slight nuttiness and umami” usually associated with meat… or, at least, that’s what research lead Jinkee Hong swears.

“We tried it with various accompaniments and it pairs well with a range of dishes,” he relayed in a Wednesday profile at The Guardian.

Hong and his collaborators have detailed their process in a new paper published with Matter. Before unleashing their Frankenstein concoction into the world, the team first slathered regular rice grains with fish gelatin and injected them with lab-grown muscle and fat stem cells. The resultant hodgepodge then cultured anywhere from 9-to-11 days before being steamed for dinner time.

[Related: Scientists made a woolly mammoth meatball.]

Depending on the meat-to-fat cell ratios, taste tests of Hong’s reportedly yielded different scent and taste palates. Higher muscular contents predictably gave hints of meat and almond, while fattier variants offered notes of cream, butter, and coconut oil. Due to the altered chemical compositions, however, the rice generally proved firmer and more brittle than standard grains. Generally, the new dish also contains 8 percent more protein and 7 percent more fat than its naturally grown source rice.

Of course, rice isn’t exactly known for its high amounts of protein or fat, so those numbers aren’t going to factor into anyone’s pre-workout meal prep anytime soon. The real benefits to such a food alternative, argues researchers, is its impressively sustainability and cost-saving potential.

By their calculations, beef rice “has a significantly smaller carbon footprint at a fraction of a price.” Real beef farming releases nearly 50 kg (110 lbs) of CO2 emissions per 100 g of protein—the hybrid grain, meanwhile, releases less than 6.27 kg (14.8 lbs) for the same amount. And while beef costs less than $14.90 per kg (2.2 lbs), the equivalent rice might only set you back $2.23.

For what it’s worth, it doesn’t sound like the mad scientists behind beef rice expect their pink granules to replace your next hot pot’s bottom layer anytime soon. Instead, such a creation could find its way into emergency food supplies in regions struck by famine or natural disaster, as well as potentially within astronaut and military rations.

“While it does not exactly replicate the taste of beef, it offers a pleasant and novel flavor experience,” Hong said. Hungry yet?

The post Scientists swear their lab-grown ‘beef rice’ tastes ‘pleasant’ appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

The next solar eclipse to cross North America is fast approaching, but over on Mars, the Red Planet already experienced one of its own celestial shadow events this year.

On February 8, the asteroid-sized Martian moon Phobos crossed in front of the sun above Jezero Crater—the area just so happening to host NASA’s Perseverance rover. As Phobos continued across the sky, Percy’s left Mastcam-Z camera angled away from its usual landscape vista subject matter towards the satellite, snapping a few dozen photos for project coordinators back at NASA’s Jet Propulsion Laboratory (JPL).

The images showcase a markedly different full lunar eclipse than the ones Earth receives every 2.5 or so years. Given both Phobos’ size and shape, the moon doesn’t fully cover the sun—instead, the  17x14x11 mile misshapen hunk of rock blocks only a small portion of the star as it continues along its path. The result arguably resembles more googly eye than awe-inspiring cosmic calendar occurrence, but it’s still a pretty impressive vantage point.

Phobos and its smaller sibling moon Deimos were discovered in 1877 by US astronomer Asaph Hall, and are respectively named after the Greek words for “Fear” and “Dread.” The origins of both satellites aren’t wholly understood, although astronomers theorize them to be either asteroids or debris leftover from the solar system’s formation that occurred around 4.5 billion years ago.

[Related: The Mars Express just got up close and personal with Phobos.]

While the Earth’s moon continues to inch away from its planetary pull at a rate of roughly 1.5 inches per year, Phobos is actually being drawn towards Mars—about six feet closer every century. While that makes for a comparatively slow descent, it does still mean the moon will eventually either crash into Mars, or break it up into thousands of fragments to form a planetary ring like Saturn’s. No need to worry, though, since that grand finale isn’t expected for another 50 million years. In the meantime, Phobos will continue orbiting Mars at a rate of three times per day, while the slower Deimos completes its journey every 30 hours.

Perseverance’s lunar eclipse capture, while incredible on its own, naturally fails to capture much detail of the moon’s pockmarked surface. Luckily, the European Space Agency’s Mars Express caught a closer look back in 2022, when the satellite came within just 52 miles of the moon to snap its own photos.

The post A Martian solar eclipse turns the sun into a giant googly eye appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

In honor of Charles Darwin’s 215th birthday on February 12, an online 300-page catalog of the famed British naturalist’s vast personal library has been released online. The Complete Work of Charles Darwin Online contains a list of the 7,400 titles and about 85 percent of the collection were previously unknown or unpublished. The items include journals, books, and pamphlets that all belonged to Darwin during his lifetime. Previously, only 15 percent of his vast collection had been readily available online, but this new collection contains 9,300 links to copies of the works all freely available. You can explore the catalog here. 

[Related: Letters From Charles Darwin.]

“This unprecedentedly detailed view of Darwin’s complete library allows one to appreciate more than ever that he was not an isolated figure working alone but an expert of his time building on the sophisticated science and studies and other knowledge of thousands of people,” John van Wyhe, a science historian at the National University of Singapore who led the project, said in a statement. “Indeed, the size and range of works in the library makes manifest the extraordinary extent of Darwin’s research into the work of others.”

Darwin’s ‘eclectic’ collection

Most of the books and writings are scientific, especially volumes on biology and geology. However, Darwin’s library also included works about his various other interests–farming, animal breeding and behavior, geographical distribution, psychology, philosophy, religion, art, history, travel, and language. Most of the material is written in English, but close to half are written in German, French, Italian, Danish, Dutch, Swedish, Spanish, and Latin.

A coffee table book published in 1872 that includes photographs of various artworks called Sun Pictures is one of the books that were not previously known to be in Darwin’s vast library. He also had a copy of Explorations and Adventures in Equatorial Africa, a science book on gorillas that was very popular after Darwin’s hallmark On the Origin of Species was published.

Among the new shorter items is an article from The American Entomologist titled “The hateful or Colorado grasshopper” and more curiously titled papers including, “The anatomy of a four-legged chicken” and “Epileptic guinea-pigs.” Darwin’s library also had a German scientific periodical published in 1877 that includes the first published photographs of bacteria. 

While several books have been written about Darwin and his work, this is the first time that his library has been completely itemized. According to Van Wyhe the catalog shows that Darwin was, “not an isolated figure working alone but an expert of his time building on the sophisticated science and studies and other knowledge of thousands of people. The size and range of works in the library makes manifest the extraordinary extent of Darwin’s research into the work of others,” Van Wyhe told The Guardian.

Mini mysteries

Darwin died in 1882 at the age of 73 and scholars believed for several his collection contained as many 1,480 books. The surviving books are in two main locations–Darwin’s former home of Down House and the University of Cambridge.

However, while a great deal of his library was preserved and cataloged, many other items were dispersed or lost. Over a period of 18 years, The Darwin Online project identified thousands of the scientist’s more obscure references in his own detailed lists of his literature. This is the first time that many of the items have even been published.

Each of Darwin’s references to a strange pamphlet or journal required solving a mini-detective story to find the publications that Darwin had referenced in his various notes. Some missing details including author, publication date, or the source of clippings in thousands of the records from older catalogs have been identified for the first time in this new online collection. 

The 426-page handwritten Catalogue of the Library of Charles Darwin helped the team reveal some of the missing documents. The catalog was compiled in 1875, and includes some of the abbreviated references to various parts of the library that scholars were able to use to track down some of the unbound pamphlets and volumes in his collection. 

[Related: Mission to recreate Darwin’s scientific Beagle voyage sets sail.]

Some of the other sources the scholars used to tease out clues include Darwin’s detailed reading notebook, Emma Darwin’s diaries, a catalog of books given to the Cambridge Botany School in 1908, and 30 volumes of correspondence. 

After Darwin’s death, all of his books were valued at only 66 pounds and 10 shillings–about £4,400 or $5,545 today. Any book belonging to or by the scientist and one-liner king is now worth an enormous sum of money to collectors. First editions of On the Origin of Species have sold for over half a million dollars and a manuscript signed by Darwin was purchased for close to $900,000 at auction in 2022.

The post Charles Darwin’s eclectic personal library is now online appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

Vandals thoroughly obliterated a Waymo autonomous taxi in San Francisco’s Chinatown on Saturday evening to the cheers of onlookers. In an emailed statement provided to PopSci, a Waymo spokesperson confirmed the vehicle was empty when the February 10 incident began just before 9PM, and no injuries were reported. Waymo says they are also “working closely with local safety officials to respond to the situation.”

A San Francisco Fire Department (SFFD) representative also told PopSci responders arrived on the scene at 9:03PM to a “reported electric autonomous vehicle on fire” in the 700 block of Jackson St., which includes a family owned musical instrument store and a pastry shop.

“SFFD responded to this like any other vehicle fire with 1 engine, 1 truck, and for this particular incident the battalion chief was on scene as well,” the representative added in their email.

Waymo Vehicle surrounded and then graffiti’d, windows were broken, and firework lit on fire inside the vehicle which ultimately caught the entire vehicle on fire. #SFFD
Photos by Séraphine Hossenlopp pic.twitter.com/aOTqL3Rk8V

— SAN FRANCISCO FIRE DEPARTMENT MEDIA (@SFFDPIO) February 11, 2024

Multiple social media posts over the weekend depict roughly a dozen people smashing the Waymo Jaguar I-Pace’s windows, covering it in spray paint, and eventually tossing a firework inside that set it ablaze—all to the enthusiastic encouragement of bystanders. After posting their own video recordings to X, one onlooker told Reuters that someone wearing a white hoodie “jumped on the hood of the car and literally WWE style K/O’ed the windshield & broke it.” Additional footage uploaded by street reporter “Franky Frisco” to their YouTube channel also shows emergency responders dousing the flaming EV, which reportedly caught fire after someone tossed a firecracker inside the car. Chinatown’s streets were already crowded by visitors attending Lunar New Year celebrations.

so um, don’t throw fireworks at waymo cars pic.twitter.com/A0OclIOJNi

— Sam Iglesias (@siglesias) February 11, 2024

Speaking to The Autopian, Frisco says that they have covered similar autonomous vehicle situations in the past, but this weekend’s drama left the Waymo vehicle looking “completely ‘decapitated.’” Upon arrival, emergency responders reportedly even had difficulty discerning whether it was a Waymo or Zoox car. Although both companies (owned by Google and Amazon, respectively) offer driverless taxi services, neither fleet resembles one another—when they are in better condition.

[Related: Self-driving taxis blocked an ambulance and the patient died, says SFFD.]

Motive for Saturday night’s incident remains unclear. The event took place as locals continue to push back against autonomous taxi operations in the area. Since receiving a regulatory greenlight for 24/7 services in August 2023, numerous reports detail cars from companies like Waymo, Zoox, and Cruise creating traffic jams, running stop signs, and blocking emergency responders. In October 2023, a Cruise driverless taxi allegedly hit a pedestrian and dragged her 20-feet down the road. Cruise’s CEO stepped down the following month, and the General Motors-owned company subsequently issued first San Francisco, then nationwide, operational moratoriums.

Not only is this weekend’s autonomous taxi butchering aggressive, dangerous, and illegal—it’s also apparently a bit of overkill. According to previous reports, driverless car protestors around San Francisco have found that simply stacking orange traffic cones atop a taxi’s hood renders its camera navigation system useless until the obstruction is removed.

The post A crowd torched a Waymo robotaxi in San Francisco appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

Damaged ancient papyrus scrolls dating back to the 1st century CE are finally being deciphered by the Vesuvius Challenge contest winners using computer vision and AI machine learning programs. The scrolls were carbonized during the eruption of Italy’s Mount Vesuvius in 79 CE and have been all-but-inaccessible using normal restoration methods, as they have been reduced to a fragile, charred log. Three winners–Luke Farritor (US), Youssef Nader (Egypt), and Julian Schilliger (Switzerland)–will split the $700,000 grand prize after deciphering roughly 2,000 characters making up 15 columns of never-before-seen Greek texts.

[Related: AI revealed the colorful first word of an ancient scroll torched by Mount Vesuvius.]

In October 2023, Farritor, a 21-year-old Nebraska native and former SpaceX intern won the challenge’s “First Word” contest after developing a machine learning model to parse out the first few characters and form the word Πορφύραc—or porphyras, ancient Greek for “purple.” He then teamed up with Nader and Schlinder to tackle the remaining fragments using their own innovative AI programs. The newly revealed text is an ancient philosopher’s meditation on life’s pleasures—and a dig on people who don’t appreciate them.  

A 1,700 year journey

The scrolls once resided within a villa library believed to belong to Julius Caesar’s father-in-law, south of Pompeii in the town of Herculaneum. Upon its eruption, Mount Vesuvius’ historic volcanic blast near-instantly torched the library before subsequently burying it in ash and pumice. The carbonized scrolls remained lost for centuries until rediscovered by a farmer in 1752. Over the next few decades, a Vatican scholar utilized an original, ingenious weighted string method to carefully “unroll” much of the collection. Even then, the monk’s process produced thousands of small, crumbled fragments which he then needed to laboriously piece back together.

Fast forward to 2019, and around 270 “Villa of the Papyri” scrolls still remained inaccessible—a lingering mystery prompting a team at the University of Kentucky to 3D scan the archive and launch the Vesuvius Challenge in 2023. After releasing open-source software alongside thousands of 3D X-ray scans made from three papyrus fragments and two scrolls, challenge sponsors offered over $1 million in various prizes to help develop new, high-tech methods for accessing the unknown contents.

What do the scrolls say?

According to a February 5 post on X from competition sponsor Nat Friedman, the first scroll’s final 15 columns were likely penned by Epicurean philosopher Philodemus, and discuss “music, food, and how to enjoy life’s pleasures.”

According to the Vesuvius Challenge announcement, two columns of the scroll, for example, center on whether or not the amount of available food influences the level of pleasure diners will feel from their meals. In this case, the scroll’s author argues it doesn’t: “[A]s too in the case of food, we do not right away believe things that are scarce to be absolutely more pleasant than those which are abundant.”

“In the closing section, he throws shade at unnamed ideological adversaries—perhaps the stoics?—who ‘have nothing to say about pleasure, either in general or in particular,'” Friedman also said on X.

Although much more remains to be uncovered, challenge organizers have previously hypothesized the scrolls could include long-lost works including the poems of Sappho.

But despite the grand prize announcement, the Vesuvius Challenge is far from finished—the newly translated text makes up just 5 percent of a single scroll, after all. In the same X announcement, Friedman revealed the competition’s next phase: a new, $100,000 prize to the first team to retrieve at least 90 percent of the four currently scanned scrolls.

At this point, learning the ancient scrolls’ contents is more a “when” than an “if” for researchers. Once that’s done, well, huge sections of the Villa of the Papyri remain unexcavated. And within those ruins? According to experts, potentially thousands more scrolls await eager eyes.

The post 2,000 new characters from burnt-up ancient Greek scroll deciphered with AI appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

In just a few short years, AI-generated deepfakes of celebrities and politicians have graduated from the confines of academic journals to trending pages on major social media sites. Misinformation experts warn these tools, when combined with strained moderation teams at social media platforms, could add a layer of chaos and confusion to an already contentious 2024 election season. 

Now, Google is officially adding itself to a rapidly growing coalition of tech and media companies working to standardize a digital badge that reveals whether or not images were created using generative AI tools. If rolled out widely, the “Content Credential” spearheaded by The Coalition for Content Provenance and Authenticity (C2PA) could help bolster consumer trust in the provenance of photos and video amid a rise in deceptive AI-generated political deepfakes spreading on the internet. Google will join the C2PA as steering member this month which puts them in the same company as Adobe, Microsoft, Intel, and the BBC. 

In an email, a Google spokerson told PopSci that the company is currently exploring ways to use the standard in its suite of products and will have more to share “in the coming months.” The spokesperson says Google is already exploring incorporating Content Credentials into the “About this image” feature in Google Image search. Google’s support of these credentials could drive up their popularity but their overall use still remains voluntarily in lieu of any binding federal deepfake legislation. That lack of consistency gives deepfake creators an advantage. 

What are Content Credentials?

The (C2PA) is a global standards body created in 2019 with the main goal of creating technical standards to certify who where and how a piece of digital content was originally created. Adobe, which led the Content Authenticity Initiative (CAI), and its partners were already concerned about the ways AI generated media could erode public trust and amplify misinformation online years before massively popular consumer generative AI tools like OpenAI’s DALL-E gained momentum.

That concern catalyzed creation of Content Credentials, a small badge companies and creators can choose to attach to an image’s metadata that discloses who created it and when the image was made. It also discloses to viewers whether or not the digital content was created using an generative AI model and even names the particular model used as well as whether or not it was digitally edited or modified later. 

Content Credential supporters argue the tool creates a “tamper-resistant metadata” record that travels with digital content and can be verified at any point along its life cycle. In practice, most users will see this “icon of transparency” pop up as a small badge with the letters “CR” appearing in the corner of the image. Microsoft, Intel, ARM, and the BBC are also all members of the C2PA steering committee.

“With digital content becoming the de facto means of communication, coupled with the rise of AI-enabled creation and editing tools, the public urgently needs transparency behind the content they encounter at home, in schools, in the workplace, wherever they are,” Adobe General Counsel and Chief Trust Officer Dana Rao said in a statement sent to PopSci. “In a world where all digital content could be fake, we need a way to prove what’s true.” 

Users who come across an image pinned with the Content Credential can click on the badge to inspect when it was created and any edits that may have occurred since then. Each new edit is then bound to the photo or video’s original manifest which travels with it across the web. 

If a reporter were to crop a photo that was previously edified using Photoshop, for example, both of those changes to the images would be noted in the final manifest. CAI says the tool won’t prevent anyone taking a screenshot of an image, however, that screenshot would not include CAI metadata from the original file, could be a hint to viewers that it was not the original file.The symbol is visible on the image but is also included in its metadata which, in theory, should prevent a trouble-maker from using Photoshop or another editing tool to remove the badge. 

If an image does not have a visible badge on it, users can copy it and upload it to this Content Credentials Verify link to inspect its credentials and see if it has been altered over time. If the media was edited without in a way that didn’t meet the C2PA’s specification during some part of its life cycle, users will see a “missing” or “incomplete” marker. The Content Credential feature dates back to 2021. Adobe has since made it available to Photoshop users and creators producing images using Adobe’s Firefly AI Image generator. Microsoft plans to use the badge with images created by its Bing AI image generators. Meta, which owns Facebook and Instagram, similarly announced it would add a new feature to let users disclose when they share AI-generated video or audio on its platforms. Meta said it would begin applying these labels “in the coming months.” 

Why Google joining C2PA matters

Google’s involvement in the C2PA is important, first and foremost, because of the search giant’s massive digital footprint online. The company is already exploring ways of using these badges across its wide range of online products and services, which notably includes YouTube. The C2PA believes Google’s participation could put the credentials in front of more eyeballs, which could drive broader awareness of the tool as an actionable way to verify digital content, especially as political deepfakes and manipulated media gain traction online. Rao described Google’s partnership as a “watershed moment” for driving awareness to Content Credentials. 

“Google’s industry expertise, deep research investments, and global reach will help us strengthen our standard to address the most pressing issues around the use of content provenance and reach even more consumers and creators everywhere,” Rao said. “With support and adoption from companies like Google, we believe Content Credentials can become what we need: a simple, harmonized, universal way to understand content.” 

The partnership comes three months after Google announced it would attach a digital watermark with SynthID to audio created using its DeepMind AI Lyrica model. In that case, DeepMind says the audio watermark shouldn’t be audible to a human ear and similarly shouldn’t disrupt a user’s listening experience. Instead, it should serve as a more transparent safeguard to protect musicals from AI generated replicas of themselves or to prove whether or not an questionable clip was genuine or AI generated. 

Deepfake-caused confusion could make already contentious 2024 elections worse 

Tech companies and media companies are rushing to establish trusted ways to verify the provenance of digital media online ahead of what misinformation experts warn could be a mind bending 2024 election cycle. Major political figures, like Republican presidential candidate Donald Trump and Florida Governor Ron DeSantis have both already used generative AI tools to attack each other. More recently in New Hampshire, AI vocal cloning technology was used to make it appear as if President Joe Biden was calling residents urging them not to vote in the January primary election. The state’s attorney general’s office has since linked robocalls to two companies based in Texas. 

But the threats extend beyond elections too. For years, researchers have warned the rampant spread of increasingly convincing AI-generated deepfake images and videos online could lead to a phenomena called the “Liar’s Dividend” where consumers doubt whether anything they see online actually as it seems. Lawyers, politicians, and police officers have already falsely claimed legitimated images and videos were AI-generated to try and win a case or seal a conviction. 

Content Credential helps could help, but they lack teeth 

Even with Google’s support, Content Credentials remain entirely voluntary. Neither Adobe nor any regulatory body are forcing tech companies or their users to dutifully add provenance credentials to their content. And even if Google and Microsoft do use these markers to disclose content made using their own particular AI generators, nothing currently stops political bad actors from cobbling together a deepfake using other open source AI tools and then try to spread it via social media.

In the US, the Biden Administration has instructed the Commerce Department to create new guidelines for AI watermarking and safety standards tech firms building generative AI models would have to adhere to. Lawmakers in Congress have proposed federal legislation requiring AI companies include identifiable watermarks on all AI-generated content, though it’s unclear whether or not that would work practically. 

Tech companies are working quickly to put in place safeguards against deepfake but with a major presidential election less than seven months away, experts agree it’s likely confusing or misleading AI material will likely play some role.

The post Google wants to fight deepfakes with a special badge appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

That one aunt of yours (you know the one) may finally think twice before forwarding Facebook posts of “lost” photos of hipster Einstein and a fashion-forward Pope Francis. On Tuesday, Meta announced that “in the coming months,” it will attempt to begin flagging all AI-generated images made using programs from major companies like Microsoft, OpenAI, Midjourney, and Google that are flooding Facebook, Instagram, and Threads. 

But to tackle rampant generative AI abuse experts are calling “the world’s biggest short-term threat,” Meta requires cooperation from every major AI company, self-reporting from its roughly 5.4 billion users, as well as currently unreleased technologies.

Nick Clegg, Meta’s President of Global Affairs, explained in his February 6 post that the policy and tech rollouts are expected to debut ahead of pivotal election seasons around the world.

“During this time, we expect to learn much more about how people are creating and sharing AI content, what sort of transparency people find most valuable, and how these technologies evolve,” Clegg says.

[Related: Why an AI image of Pope Francis in a fly jacket stirred up the internet.]

Meta nebulous roadmap centers on working with “other companies in [its] industry” to develop and implement common identification technical standards for AI imagery. Examples might include digital signature algorithms and cryptographic information “manifests,” as suggested by the Coalition for Content Provenance and Authenticity (C2PA) and the International Press Telecommunications Council (IPTC). Once AI companies begin using these watermarks, Meta will begin labeling content accordingly using “classifiers” to help automatically detect AI-generated content.

“If AI companies begin using watermarks” might be more accurate. While the company’s own Meta AI feature already labels its content with an “Imagined with AI” watermark, such easy identifiers aren’t currently uniform across AI programs from Google, OpenAI, Microsoft, Adobe, Midjourney, Shutterstock, and others.

This, of course, will do little to deter bad actors’ use of third-party programs, often to extremely distasteful effects. Last month, for example, AI-generated pornographic images involving Taylor Swift were shared tens of millions of times across social media.

Meta made clear in Tuesday’s post these safeguards will be limited to static images. But according to Clegg, anyone concerned by this ahead of a high-stakes US presidential election should take it up with other AI companies, not Meta. Although some companies are beginning to include identifiers in their image generators, “they haven’t started including them in AI tools that generate audio and video at the same scale, so we can’t yet detect those signals and label this content from other companies,” he writes.

While “the industry works towards this capability,” Meta appears to shift the onus onto its users. Another forthcoming feature will soon allow people to disclose their AI-generated video and audio uploads—something Clegg may eventually be a requirement punishable with “penalties.”

For what it’s worth, Meta also at least admitted it’s currently impossible to flag all AI-generated content, and there remain “ways that people can strip out invisible markers.” To potentially address these issues, however, Meta hopes to fight AI with AI. Although AI technology has long aided Meta’s policy enforcement, its use of generative AI for this “has been limited,” says Clegg, “But we’re optimistic that generative AI could help us take down harmful content faster and more accurately.”

“While this is not a perfect answer, we did not want to let perfect be the enemy of the good,” Clegg continued.

The post Sharing AI-generated images on Facebook might get harder… eventually appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

Bluesky is finally open to anyone still desperate to fill the social media void left by Twitter’s collapse into the black hole that is X. Although Twitter founder and former CEO Jack Dorsey first announced the “decentralized” platform back in 2019 (more on that later), the project truly launched in 2021. After debuting on iOS in February 2023, it remained an invite-only, waitlisted site for months. As of February 6, however, anyone can now create a free account on the site.

[Related: Twitter alternative Bluesky is fun, friendly, and kind of empty.]

Using Bluesky looks and feels much like Twitter before Elon Musk pulled off his tumultuous, $44 billion company takeover in 2022. The focus remains primarily on microblogging through 300-character limited text and image posts, all displayed in chronological timelines against a very similar color palette. There were an estimated 3 million accounts on Bluesky prior to today’s public launch, which currently makes for a far quieter social media experience when compared to the numbers seen on established platforms like X (397 million users), Facebook (2.9 billion), TikTok (1 billion), and even Threads (130 million), Meta’s recent attempt at a Twitter replacement. That, of course, may change in the coming weeks.

The biggest differences between Bluesky and its competitors reside within the website’s inner workings. Its underlying technology is available as an open, decentralized protocol, meaning any motivated developers can access the code to create their own algorithmic tweaks and, eventually, even spin-off social networks. Social media alternatives like Mastodon also operate using an open, decentralized, protocol, although the nitty-gritty is still confusing years after its debut. (TechCrunch has a solid rundown of protocol jargon and its uses.)

Although Dorsey originally intended to transition Twitter over to this open protocol format, he later altered plans to make Bluesky a wholly separate project. After Musk’s 2022 takeover, many users decided to abandon his newly christened X due to its rapid devolution into an even more unreliable and toxic ecosystem. As the social media landscape continued to shift, Bluesky’s closed-door era offered marginalized communities a relatively reliable, supportive online safe haven.

Bluesky’s developers gradually opened the site to more users throughout the last year, but not without issues. Many users continue to voice concerns regarding an uptick in the kinds of harassment that forced them from X in the first place. With only a couple dozen employees, it remains to be seen if Bluesky can handle a deluge of additional users. In its public launch announcement, the company’s developer team says it plans to focus on improved safety, fact-checking, and content moderation policies in the coming weeks.

The post Twitter alternative Bluesky is now open to the public appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

Over a hundred new emojis are on deck to arrive in the coming weeks, including six brand new designs such as a fiery phoenix, a lime wedge, a broken chain, and a (nonpoisonous!) mushroom and an overhaul of all family options. Selected from the Unicode Consortium’s September 2023 recommendations, the latest additions are currently only available for developers in iOS 17.4 beta 1, but are expected to issue with the final iOS 17.4 public release, likely sometime in March or April 2024, according to Emojipedia.

[Related: Meet the newest Apple emojis: a goose, a moose, and another pink heart.]

A couple of extra smiley faces will find their way onto users’ devices this spring—one indicating an affirmative, up-and-down headshake, while the other conveys a negative, side-to-side motion. According to Emojipedia, however, the majority of the latest batch focuses on “new direction-specifying versions of six existing people emojis” which tally up to 108 emojis once all five skin tones and three gender options are accounted for.

It’s been nearly a year since users last saw a new class of emojis. In February 2023, Emojipedia confirmed a moose, a ginger root, and a goose, among many others. As is the case with every such announcement, the latest designs could see some final tweaks before their public release later this spring—such as past alterations for the troll and peach emojis.

Perhaps the most noticeable changes are the revamped family designs, which replace their more detailed, colorful renderings with white silhouettes overlaid atop a bluish slate square icon. While not as precise as the current iterations, removing minutiae like hair styles, colors, and facial shapes should both make the options more inclusive, as well as simpler to select from for users.

How are emojis chosen?

The process for determining new emojis is surprisingly complex. A potential addition’s journey begins with the Unicode Consortium, an international body tasked with helping oversee the standardization and implementation of multilingual text and imagery updates. Anyone can submit their own original emoji ideas following these guidelines, which Unicode then reviews and narrows down the options. From there, card-carrying consortium members (who annually shell out $21,000 for the privilege) vote on a final suite of emojis, which then move on to final designing and integration before arriving in regularly scheduled software updates.

While you may not get the chance to strut your stuff with a phoenix text until later this spring, impatient emoji aficionados can look ahead to what may come next to their keyboards. According to Emojipedia, you could soon see a shovel, a leafless tree, as well as a likely very relatable “Face with Bags Under Eyes.”

The post The latest emojis feature a mushroom, a phoenix, a lime, and more inclusive family options appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

Videos of what looks like Tesla drivers using the new Apple Vision Pro “spatial computing” headset while in Autopilot mode are going viral, but at least one is staged. After getting over 24 million views on X, 21-year-old Dante Lentini may still face legal repercussions for his stunt.

In an email to PopSci on Monday, Lentini confirmed a video appearing to show him being stopped by police for using Apple’s $3,499 headset behind the wheel of his Tesla was filmed in a “skit-style fashion.” The 25-second clip shows Lentini sitting in the Tesla driver’s seat while traveling on a highway using Autopilot. Instead of keeping his hands on the steering wheel, as Tesla directs all users to do while in Autopilot, Lentini gestures to imply he is using Vision Pro’s interface. (The Apple headset relies on interpreting specific hand movements to navigate and utilize its apps.) The video then cuts to Lentini in a parking lot as a police vehicle flashes its lights behind him.

Think different. #applevisionpro pic.twitter.com/dEALUsntS8

— Dante (@lentinidante) February 2, 2024

“So the police were not even in the parking lot for me to begin with,” Lentini alleges in the email. “I wasn’t pulled over never mind [sic] not being arrested nor ticketed.”

Lentini uploaded his clip to X on February 2, the same day Apple’s Vision Pro headset hit stores, but it wasn’t until this weekend that the post began gaining momentum. Numerous outlets have since covered Lentini’s video, as well as similar content. A different video posted to X on February 3 appears to show another Apple Vision Pro user in the driver’s seat of a Tesla Cybertruck. Like Lentini, the driver makes gestures known to control the headset, implying the $60,990 base price EV is engaged in Autopilot or Full Self-Driving Beta mode. The Cybertruck video has racked up over 17 million views by Monday morning.

In a follow-up email to PopSci, Lentini confirmed he used Tesla’s Autopilot program during his video after he “got over to the right most lane [sic].” He also claimed he only wore Apple’s headset for “10-15 second increments” totalling “less than 30-40 seconds combined.” 

“I believe the Vision Pro doesn’t even work while traveling since the technology fails to be able to track your reference surroundings and place the graphics accordingly,” he continued. “So all it showed was a pass through video feed,” referring to the headset’s ability to visualize external surroundings with a reportedly 12 millisecond latency, “as if I was just wearing sunglasses.”

[Related: Here’s a look at Apple’s first augmented reality headset.]

Most US state traffic laws prohibit wearing anything that could potentially obscure a driver’s ability to see their surroundings. In Palo Alto, where Lentini claims to reside, “it is unlawful for a person to drive a vehicle if a television receiver, a video monitor, or a television or video screen, is operating and is visible to the driver.” Violations could include a fine of $238, as well as a point added to the driver’s DMV record.

A previous review of the parameters within Vision Pro’s visionOS coding indicates it disables certain features if it detects users traveling over a “safe speed,” although it’s unclear if this applies to driving. A separate “Travel Mode” can reportedly be enabled while “stationary” in an airplane, but Apple does not offer an explanation of how Vision Pro assesses the speed, travel, and passenger status. 

According to Apple’s official product page, the Vision Pro includes built-in safety features meant to help prevent collisions and falls. “[I]t’s also important to use the device in a safe manner. For example, don’t run while wearing Apple Vision Pro, use it while operating a moving vehicle, or use it while intoxicated or otherwise impaired,” the company states.

Lentini suspects similar viral content videos are also “skits.” Although he understands “some people’s initial frustration” after seeing his clip, “there’s nothing obstructing my vision. I personally feel like it’s more dangerous to text and drive or even eat and drive, even though I still recommend not wearing these while driving.” Illegal “distracted driving” is defined on a state-by-state basis, but usually includes texting. In some places, eating can also fall within the bounds of distracted driving. 

Whether or not flashy, bank-draining luxury items like Apple Vision Pro and Tesla Cybertruck will prove successful remains to be seen. For now, at least, the combination is leaving bystanders dizzied by the whirlwind mix of legality, wealth, virality, and veracity—all exacerbated by such posts’ ability to spread across platforms like X.

The post Don’t worry, that Tesla driver only wore the Apple Vision Pro for ’30-40 seconds’ appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

Contrary to what many consumers could be forgiven for thinking, Comcast does not offer 10G cable internet connections—at least, not in the way you think. Technically, ’10G internet’ does not exist yet, but that didn’t stop Comcast from advertising a product that could easily be confused as such for months. Following a decision from the National Advertising Review Board earlier this week, however, the telecom company will need to be much clearer when advertising their internet service options. 

On January 31, an NARB panel determined Comcast’s “Xfinity 10G Network” branding campaign “expressly communicates at a minimum that users of the Xfinity network will experience significantly faster speeds than are available on 5G networks.” This, however, is often far from the case. In fact, as Ars Technica explained on Wednesday, the only way customers could obtain internet speeds close to what Comcast advertised is through its “Gigabit Pro” fiber-to-home connection package costing $299.95 per month atop a $19.95 modem lease fee. That’s after you pony up Gigabit Pro’s $500 installation charge and a $500 activation fee, mind you. For reference, the median Xfinity user download speeds currently clock in around 236Mbs—not bad, but certainly not 10Gbs.

[Related: Read this before setting up a 5G network at home.]

Incidentally, “gigabit” is the key term in all this confusion. Whereas the “G” in a phrase like “5G wireless connection” refers to the technology’s fifth generation, Comcast’s Xfinity 10G signifies its (often aspirational) 10-gigabit broadband internet speed package. 10G wireless connections—the kind that T-Mobile and Verizon claimed Comcast deliberately tried to confuse customers about—aren’t on the market.

The first instances of nebulous 10G marketing campaigns date back to a 2019 announcement from prominent industry trade group, the National Cable & Telecommunications Association (NCTA). The reveal frustrated many at the time, given that even 1Gps speeds remained unavailable to most consumers. In February 2023, however, Comcast embraced the terminology for its fixed internet network system, Xfinity 10G Network. This decision finally prompted T-Mobile and Verizon to file a complaint with the National Advertising Division (NAD), an industry watchdog, which subsequently recommended Comcast abandon its 10G branding efforts in October 2023. Comcast unsurprisingly appealed the decision, but this week’s new NARB ruling finally forced the company’s hand.

“Although Comcast strongly disagrees with NARB’s analysis and approach, Comcast will discontinue use of the brand name ‘Xfinity 10G Network’ and will not use the term ‘10G’ in a manner that misleadingly describes the Xfinity network itself,” the company said in a statement alongside the NARB’s January 31 announcement. It’s unclear what the new branding will be, but it certainly seems unlikely “10G” will be a part of it.

But because the NARB ruling isn’t technically a wholesale ban on using “10G” in advertising or products, don’t expect the terminology to completely disappear. In fact, Comcast already made it pretty clear it intends to continue wedging 10G(igabit) phrasing into ads wherever it can. According to Engadget, the company still “reserves the right” to keep using “10G” and “Xfinity 10G,” so long as it doesn’t “misleadingly describe the Xfinity network itself.”

Meanwhile, the NCTA just so happened to celebrate the 5-year anniversary of its original 10G campaign rollout… lest you think confusing telecom advertising is truly disappearing anytime soon.

The post No, ’10G internet’ is not a thing appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

It’s relatively easy to trace a written linguistic history—there’s generally a lot of written documentation and records to study. Things get trickier, however, when attempting to examine a sign language’s evolution. Most transformations within the currently over 300 known sign languages (or SLs) around the world occurred sans text over generations of learners. Add in the centuries of marginalization experienced by Deaf and hard of hearing communities, and establishing concrete relationships between SLs becomes extremely difficult.

To help correct this long standing issue, researchers recently created a novel computational program capable of analyzing the relationships between various SLs. The result, published today in Science, is a first-of-its-kind large-scale study that greatly expands on linguists’ understanding of sign language development while challenging long held beliefs about its evolution.

[Related: Online classes are difficult for the hard of hearing. Here’s how to fix that.]

“Many people mistakenly think that sign language is shared around the world, but really the world is full of a vibrant tapestry of different sign languages,” Natasha Abner, study lead author and an associate professor of linguistics at the University of Michigan, writes in an email to PopSci.

For their study, Abner and her colleagues first compiled a video dictionary of core, “resilient” vocabulary across 19 modern sign languages, such as American, British, Chinese, French, Japanese, and Spanish, among others. For example, while a sign for “oak tree” may only occur in languages spoken in regions with oak trees, the concept of just a “tree” is much more ubiquitous. Researchers then broke down video demonstrations for the 19 signing variants for “tree” (along with many other words) into basic phonetic parameters, then entered it all into a massive database.

“What we do in the study is look at how the sign languages refer to these commonplace, universal objects in the world and we work backwards to build a history of the language and languages,” Abner says. “This built history helps us understand the histories of the communities in ways that the historical records cannot because they are so limited and sparse.”

The computational analysis program then examined the signed vocabulary glossary, categorizing each entry based on intricate factors like handedness (one- or two-handed signs), handshape, location, and movement.

“This coding system avoids outcomes driven by superficial similarities or differences in two key ways,” reads a portion of the team’s study. “One, possible character values in the coding system range from two distinct values (handedness) to 10 distinct values (handshape), so it is a highly articulated system capable of capturing and tracking fine-grained differences.”

In tracing signed vocabularies’ evolutions, researchers applied phylogenetic analysis typically associated with biologically inherited traits to physically conveyed communications.

“In our study, the ‘genes’ of language are the words that the languages use to describe the world around them,” says Abner. Pursuing this strategy meant that, instead of simply applying existing computational methods to sign language data, Abner’s team used sign languages “as the empirical basis for advancing the computational methods themselves.”

[Related: The language you speak changes your perception of time.]

After examining the dataset, the team’s program established two wholly independent European and Asian sign language families alongside family trees for each one, as well as two distinct Asian sign language subfamilies. Some of the findings reinforced the already known, lasting effects of Western colonization, such as the relationship between British, Australian, and New Zealand sign languages at the expense of endangered or extinct indigenous variants. 

Meanwhile, the documented influence of French sign language within the Western European language tree is backed up by France’s help in expanding Deaf education schools during the 18th century. At the same time, the new computational analysis also revealed previously undocumented connections between British Sign Languages and Western European varieties. To back up the program’s claim, Abner’s team referred back to limited historical records, and found them to corroborate these links.

Abner believes these findings, alongside future advances, will allow sign language linguists the ability to study even more languages and Deaf communities.

“We view this as an important component of demonstrating the equity between signed and spoken languages, and the fact that both are rooted in the biological capacity for language that is part of what makes us human,” she tells PopSci.

“If we want to understand our humanity, then we cannot limit ourselves to spoken languages.”

The post Computer modeling is tracing the hidden evolution of sign languages appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

It’s a programming question nearly as old as the inspiration itself: “Can it run Doom?”

Released over 30 years ago, the seminal first-person shooter (FPS) is a touchstone classic of video gaming—not only for its influence on the medium, but for inspiring some pretty wild coding schemes. In 1997, Doom’s creators at iD Software released the game’s original source code for free online, allowing legions of hobbyists the ability to tinker and experiment to their demonslaying heart’s content.

[Related: Using ‘Doom’ to design a room.]

Since then, enthusiasts have hacked an ever-expanding list of devices to play the pixelated sci-fi horror adventure. From pregnancy tests, to tractors, to ATMs, to calculators—if it’s got at least some circuitry in it, chances are that, with a little ingenuity, it can run Doom.

…And now, “circuitry” apparently extends to gut biome bacteria.

Lauren “Ren” Ramlan, an MIT biotechnology PhD student researcher, recently released a paper documenting a truly stomach-churning feat: As highlighted by RockPaperShotgun, she programmed Doom to run on a display made from E. coli cells.

“To run Doom, all one needs is a screen and willpower,” Ramlan writes. “… Ultimately, this begs the question of how biological systems might be engineered to host this classic millennial FPS.”

According to Ramlan, for the project to work, E. coli cells must function as traditional pixels capable of being either “on” or “off.” They also need to collectively light up to form images like a computer monitor or TV screen. To make that happen, Ramlan first grew cells within a 32×48 1-bit well plate, then connected the makeshift screen to a controller capable of processing and translating binary code into the “addition or omission of a repressor controlling the fluorescence of the cells.” Basically, Ramlan swapped a traditional screen’s tiny light diodes for glowing bacterial cells.

While likely a tall endeavor to the average layperson, Ramlan made it happen by combining that aforementioned willpower alongside some dizzying coding and organic chemistry skills. The result is a functioning display screen of luminescent E. coli capable of showing off Doom in real time… sort of.

Ramlan’s invention reportedly takes roughly 70 minutes to fully illuminate, then another 8 hours and 20 minutes to dim back to its original state. All told, that’s basically about 9 hours to offer players a single frame of the video game. Given that the original Doom tops out at 35 frames per second, it would take quite a while to actually play through the entire game—like, 599 years “a while,” according to Ramlan’s calculations.

“This is an amazing find, because it means we are a small handful of generations away from the peak of human engineering… where Doom and life become one,” she deadpans in her explainer video.

It may not be the most efficient way to stave off the demon hordes of Mars, but it certainly is one of the more creative ones yet.

The post Running ‘Doom’ on E. coli cells… very, very slowly appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

Elon Musk recently claimed that the first human patient has received a Neuralink brain-computer interface (BCI) implant, but stopped short of offering any substantial additional information or proof. In a series of messages posted on Monday evening to X, his social media platform formerly known as Twitter, Neuralink’s majority owner alleged an anonymous volunteer underwent the company’s experimental procedure on January 28 and “is recovering well” while showing “promising neuron spike detection.” In a separate tweet, Musk also announced the first Neuralink product will be called Telepathy.

Despite the potential company milestone, no other official details have been made available in the nearly 24 hours since Musk’s announcement. At the time of writing, Neuralink’s most recent news update is still an initial solicitation on their official website for human trial volunteers published last fall.

[Related: Neuralink is searching for its first human test subjects.]

Neuralink implant plans, explained 

Co-founded by Musk in 2016, Neuralink is a tech startup aiming to create BCI implants for a variety of uses. However, its first devices are specifically tailored for physically impaired and paralyzed patients suffering from diseases such ALS to restore motor functions and the ability to communicate. 

Musk, in turn, has repeatedly voiced hopes of designing implants capable of connecting anyone’s brain to smartphones, computers, and the internet. The devices are intended to be completely reversible and ungradable, with Musk likening them to iPhones. In 2022, Neuralink’s “N1” was reportedly the size of a quarter, and employed a robotic surgeon machine to subcutaneously connect microscopic wiring to the brain. Musk has also suggested users could eventually summon Teslas using their Neuralink implants, and stated he plans to eventually receive the implant.

Why it took Neuralink so long to begin human trials

Although the FDA approved Neuralink to begin human trials in September 2023, regulators initially rejected the medical startup’s application earlier that year due to multiple safety concerns. Numerous bombshell reports have accused Musk’s company of conducting horrific “hack job” surgeries, some of which resulted in severe injuries, psychological damage, and the death of over 1,500 lab test animals. A graphic exposé from Wired, for example, described dozens of botched spinal surgeries, severe self-harm behaviors, and fatal infections as a result of sheep, monkeys, and pigs receiving experimental implants. These investigations subsequently prompted multiple ongoing federal probes regarding Neuralink’s potential animal abuse, and improper transportation of hazardous materials.

Neuralink had already fallen far behind Musk’s ambitious timeline goals before word spread of the company’s allegedly troubling strategies. In 2019, for example, the SpaceX and Tesla CEO expressed a desire to receive an FDA greenlight by the end of 2020, only to later blow past a revised 2022 deadline.

Neuralink isn’t the first company to implant a BCI

If Sunday’s surgery is confirmed outside of Musk’s X posts, Neuralink will only be the latest company to successfully demonstrate human-BCI devices. Guger Technologies unveiled the world’s first BCI in 2010, while numerous other designs have since debuted, including Synchron’s “Stentrode” neuroprosthesis and one from BrainGate allowing a paralyzed man to convert imagined handwriting into text.

“Imagine if Stephen Hawking could communicate faster than a speed typist or auctioneer. That is the goal,” Musk tweeted last night.

[Related: Neuralink’s human trials volunteers ‘should have serious concerns,’ say medical experts.]

Musk’s nebulous Monday tweets recall his reportedly impulsive social media habits, a pattern that has earned its share of regulatory fallout. Problematic tweets from Musk in 2018 eventually resulted in an SEC security fraud settlement that included relinquishing his position as Tesla chairman, alongside agreeing to separate $20 million fines levied against him and his company. He is currently appealing the decision, as well as a National Labor Relations Board’s ruling pertaining to alleged union busting threats posted by Musk to Twitter.

It’s unclear why Neuralink has not released any more information supporting Musk’s tweets. Company representatives did not respond to PopSci at the time of writing.

The post Elon Musk alleges Neuralink completed its first human trial implant appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

The podcasters behind “George Carlin: I’m Glad I’m Dead”—a controversial stand-up “special” originally advertised as AI-generated—confirm their stunt routine was “completely written” by a human. Although an unsurprising turn of events, it still may not shield them from legal fury.

A brief catchup on the Carlin controversy

To bring anyone blessedly unaware of recent events up to speed: Earlier this month, content creators Will Sasso and Chad Kultgen hyped a forthcoming, Carlin-centric episode of Dudesey, a podcast series they claim is written by a “state of the art entertainment AI” of the same name trained on data including the duo’s own social media posts, text messages, and emails. Then on January 9, Sasso and Kultgen released the episode (currently private on YouTube) after “training” “AI”  (they claimed) on text and audio from the entirety of Carlin’s over 50-year career.

“George Carlin died… before 2010, I think—and now he’s been resurrected by an AI to create more material,” Kultgen said in a preview YouTube video. Carlin died in 2008.

At the episode’s outset, Dudesey “AI” claimed: “I listened to all of George Carlin’s material and did my best to imitate his voice, cadence and attitude, as well as the subject matter I think would have interested him today,” before launching into “George Carlin: I’m Glad I’m Dead.” Over the course of the segment, a vocal clone of the late comedian covered a range of Carlinesque topics, including gun violence, politics, free speech, and class.

“If you’re in America, you’re special. God made something just for you, something no other country on the planet gets,” the fake Carlin states early in the episode, as reported over the weekend by The Washington Post. “Of course, I’m talking about mass shootings!” Listeners were not amused.

A tough crowd

Virtually the only positive response to Dudesey’s fake Carlin set came from a self-provided audience laugh track. The internet quickly panned the episode as a clickbait cash-in meant to leverage a simultaneously hyped and maligned AI industry.

“ChatGPT and other LLMs rely on vast swaths of copyrighted material created by human hands. Dudesy can’t fart out a crass imitation of George Carlin without viewing 14 standup specials that are the sum of a human’s life, dreams, and labor,” Matthew Gault wrote for Vice.

Others doubted how much AI technology was actually used to make “I’m Glad I’m Dead.” Images in the YouTube video resembled generative AI artwork and vocal cloning can already produce near-indistinguishable imitations of real human voices. However, critics were skeptical that any generative AI is currently capable of creating an hour’s worth of coherent material.

“Despite the claims that Dudesy has somehow ingested Sasso and Kultgen’s work, or that it somehow ‘learns’ and ‘generates data that will be used to make the next episode better,’ it appears to be more likely that it uses a combination of readily-available tools patched together to ‘surprise’ two comedians clearly in on the act,” commentator Ed Zitron wrote in post for his internet culture newsletter.

“It’s also worth remembering the context around AI at the time Dudesy premiered in March 2022. The ‘state of the art’ public AI at the time was the text-davinci-002 version of GPT-3, an impressive-for-its-day model that nonetheless still utterly failed at many simple tasks,” Kyle Orland explained for Ars Technica. “It wouldn’t be until months later that a model update gave GPT-3 now-basic capabilities like generating rhyming poetry.”

Meanwhile, the comedy legend’s daughter also made her own thoughts on the matter clear.

“I understand and share the desire for more George Carlin. I, too, want more time with my father,” Kelly Carlin wrote in a statement posted to X a day after the video’s release. “But… the ‘George Carlin’ in that video is not the beautiful human who defined his generation and raised me with love. It is a poorly-executed facsimile cobbled together by unscrupulous individuals to capitalize on the extraordinary goodwill my father established with his adoring fan base.”

Legal fallout

The Carlin estate’s legal team filed a lawsuit against Dudesey’s creators on January 25, claiming copyright infringement, deprivation of rights of publicity, and violation of rights of publicity. According to US law, plaintiffs could be entitled to as much as $150,000 per charge. Soon afterwards, the podcasters finally confirmed many critics’ suspicions.

In a statement first provided to The New York Times on Friday morning last week, a spokesperson for Sasso and Kultgen stated their Dudesey is a “fictional podcast character created by two human beings.” As for “I’m Glad I’m Dead,” the material itself was “completely written” by Kultgen, although the lawsuit’s defendants have yet to confirm if they employed AI for the Carlin vocal clone or accompanying artwork. 

Joshua Schiller, a partner at Boies Schiller Flexner, LLP, and an attorney for the Carlin estate, believes Sasso and Kultgen admitting to the stunt won’t absolve the duo of legal responsibility.

“Who knows what to believe from these defendants? All we know is that they are craven opportunists who have fabricated a piece of content that violates multiple of my clients’ rights,” Schiller said in a statement provided to PopSci on Monday. “We look forward to getting the truth about how this shameful spectacle was created and holding defendants accountable for their blatant disregard for the law and basic decency.”

According to the lawsuit filing previously obtained by Ars Technica, plaintiff attorneys argue Carlin’s reputation and legacy is now potentially damaged by association with the Dudesey special, and are continuing to seek legal and financial compensation.

The Carlin “stand-up,” although largely debunked, draws attention once again to the mounting copyright-related lawsuits against a still largely unregulated AI industry. Makers of programs such as ChatGPT maintain that access to copyrighted material is key to training trustworthy, safe AI. Compensation for such access, however, is currently far from uniform, reliable, or even legally sound. Meanwhile, there remains the possibility “I’m Glad I’m Dead” employed both AI vocal clone and generative art programs—both of which are often trained on massive, copyrighted datasets.

The real George Carlin saw it coming

As the controversies continue to play out, it certainly feels like Carlin himself was onto something almost exactly 20 years ago.

“I’ve been uplinked and downloaded. I’ve been inputted and outsourced. I know the upside of downsizing; I know the downside of upgrading,” he wrote in his 2004 essay, “Ode to the Modern Man.” “I’m a high-tech lowlife. A cutting-edge, state-of-the-art, bicoastal multitasker, and I can give you a gigabyte in a nanosecond.”

UPDATE 04/03/2024 10:31AM: The legal team for Carlin’s estate announced an out-of-court settlement with the creators of George Carlin: I’m Glad I’m Dead. Sasso and Kultgen agreed to permanently remove the special from all platforms, and would not use Carlin’s image, voice, or likeness again with estate approval. Additional settlement details, including monetary compensation, were not disclosed.

The post Humans actually wrote that fake George Carlin ‘AI’ standup routine appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

As impressive as many biohybrid robotic projects are, they aren’t exactly known for their hairpin turns. In fact, it’s still pretty difficult to design an agile machine merging artificial materials and biological tissue. But if a future generation of biohybrids does manage to one day clear that hurdle, they could owe it to a tiny pair of cute, albeit pretty much useless, robo-legs.

Researchers at the University of Tokyo detailed their 3cm tall creation in a new study published today in Matter. By combining 3D-printed parts, rubber, and lab-cultivated rat muscle tissue cells, the team managed to create a proof-of-concept minibot capable of turning on a 90-degree pivot while suspended in water. To make it work, one “leg” receives minute electrical pulses that in turn contract its rat muscle actuators, while the other serves as its fixed point of support. In doing so, the biohybrid prototype manages to pivot at an angle previously unobtainable by similar robotic designs.

[Related: Meet xenobots, tiny machines made out of living parts.]

It’s a pretty big deal… although an incredibly slow one. According to researchers, their robot moves at an incremental 5.4mm per minute thanks to electrical stimulations issued through the water at five second intervals. But before you think this is being a bit harsh on the little guy, take it from team member Shoji Takeuchi:

“This is still basic research. We are not at the stage where this robot itself can be used anywhere,” he stipulated while speaking to New Scientist.

As it stands (so to speak), the biohybrid can’t even remain upright underwater without a buoy support system. It also needs constant supervision, and a watery conduit to stimulate the muscle actuators. Takeuchi says getting it onto dry land would require much thicker muscle designs, additional joints, and some kind of nutrient system to keep the tissue cultures alive and kicking.

Writing in their paper, researchers believe their advancements potentially could “contribute to a deeper understanding of biological locomotion mechanisms,” as well as possibly “pave the way further mimicking the intricacies of the human gait mechanism” in biohybrid robots.

After a few more years pumping weights at the gym on dry land (i.e. advancements in the lab), more complex robot iterations could possibly find their way back into water as deep-sea explorers. Science also notes biohybrid designs may also eventually be deployed in search-and-rescue missions. It may sound somewhat spooky to find yourself saved by a biorobot built from rat muscles—but it’s either that or the spider bots.

The post Watch this cool, useless biohybrid robot take a stroll appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

It’s finally arriving, but don’t panic. You’ve had plenty of time to steel yourself. After 15 years of loyal service, Microsoft Office 365’s default font is no longer Calibri. Instead, Word, Outlook, PowerPoint, and Excel users on TikTok are noticing (or will soon enough) their trusted sans-serif typeface is officially swapped for a brash upstart known as Aptos.

The change-of-the-guard began rolling out late last year through Beta and “other preview channels.” Because such updates are often staggered, however, The Verge notes many users are only recently seeing the new font default—Microsoft only created its official Aptos documentation on January 19. While originally confirmed a little over six months ago, company plans for a replacement font actually began as far back as 2021.

Such decisions aren’t arbitrary. In fact, there’s a pretty succinct reason as to why Calibri became yesterday’s font: It’s all about resolution.

It’s no coincidence that Calibri first replaced Times New Roman within Office 2007, right about the time of Apple’s iPhone debut. Back then, OLED—much less Retina, LED, or any other high-definition displays—weren’t a widespread thing. On top of lower resolutions, most computer screens (at least desktops) still measured 1024×768 or 1280×800. Still, times needed changing—Times New Roman, more specifically.

[Related: Why the State Department is the new ambassador for the Calibri font.]

After serving its purpose for years, Microsoft swapped the iconically 1990’s default for a newly designing, then-modernized font Calibri. Fast forward through roughly another decade-and-a-half of digital improvements and screen resolution advancements, and the company decided it was time to start finding its newest Font-in-Chief.

In April 2021, Microsoft’s design department announced the search was on for its next default typeface, one that could hold its own against in a world of 4K, ultra-HD, and all manner of other modern, higher-resolution options to come. Five potential replacements debuted as test options for Office users, all with very font-esque names like Grandview, Seaford, Skeena, Tenorite, and Bierstadt.

Two years later, Microsoft finally announced a decision: a slightly modified version of Bierstadt, dubbed Aptos. Described as Calibri’s “modern successor,” Aptos is meant as “the perfect font for higher resolution screens,” possessing a “sharpness [and] uniformity” with higher resolution screens in mind.

In some ways, Aptos is actually a return to form for Microsoft Office. As Ars Technica noted last year, the font was designed by world renowned type designer Steve Matteson, the man responsible for Windows 3.1’s original TrueType fonts including Arial, Courier New, and yes, Times New Roman. He also came up with Segoe, the default Windows system font since Vista, as well as the company’s current logo typeface.

“[Matteson] designed the font with a slight humanist touch,” Microsoft explained in its original Aptos reveal post. “He wanted Aptos to have the universal appeal of the late NPR newscaster Car Kasell, and the astute tone of The Late Show host Stephen Colbert.”

… So. There’s that. In any case, Aptos will now begin the de facto font option across Microsoft Office for the foreseeable future. Calibri loyalists can still revert back to their beloved design by following these easy steps, and anyone still miffed about Microsoft passing up on Grandview, Seaford, Tenorite, or Skeena contenders can still choose those, as they’re now included in Office’s already sizeable list of font options.

But while Aptos is now the default font for most Windows Office users, it’s unclear if the upper echelons of the US government will abide by the decision anytime soon. After all, the State Department just ordered employees to swap Times New Roman for Calibri in all requested documents submitted to the Executive Secretariat in February 2023.

You can read Microsoft’s full Aptos rundown here, as well as see exactly what it looks like here (kind of like Calibri).

The post Get ready for Aptos, Microsoft’s new default font appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

Meta continues to steadily roll out updates for younger users in an attempt to bolster their safety and privacy. On Thursday, the tech company announced some of its most restrictive measures yet—in theory. Teens users, by default, will no longer receive direct messages on Instagram and Facebook from anyone that isn’t a follower or connection. “Connections,” according to Meta, are those people that users have “communicated with” in some way, such as sending text messages, voice or video calls, or accepting message requests. A similar update is also going into effect on Facebook, with messages only allowed from friends and “people they’re connected to through phone contacts, for example.”

Instagram previously restricted anyone over 18 years old from messaging younger accounts that did not already follow them back. The expanded rules will automatically apply to global users under the age of either 16 or 18, depending on their country’s laws, who now also cannot message other teens they are not connected to. Similarly, group chats with teens can only include their friends or connections, and the same messaging restrictions apply for teens who don’t follow each other.

[Related: Instagram will start telling teens to put down their phones and go to sleep.]

To disable the setting, teens will need to receive permission from their parents through the social media platforms’ parental supervision tools. Until now, parents and guardians would receive notifications if teens changed their settings, but couldn’t do anything about it. According to Meta, affected users will receive a notification on their apps regarding the new changes.

“As with all our parental supervision tools, this new feature is intended to help facilitate offline conversations between parents and their teens, as they navigate their online lives together and decide what’s best for them and their family,” Meta wrote in today’s newsroom post.

It’s worth bearing in mind here that the updates assume that the parental supervision option is enabled, users have accurately entered their “declared age” on either Instagram or Facebook, and Meta’s age-predicting technology is working as planned.

The direct message changes arrive following multiple recent changes tailored for Facebook, Instagram, and Messenger’s under-18 crowds. Last week, Meta announced a new “nighttime nudge” feature that will begin politely reminding teens at regular intervals after 10pm to drop their phones and turn in for the night. Earlier this month, the company also revealed plans to roll out automatically restrictive content settings focused on curtailing young people’s exposure to potentially harmful subject matter, particularly posts and messages related to self-harm, eating disorders, and graphic violence. Unlike today’s new features, however, those content censors are mandatory, and unable to be circumvented for any accounts under the age of 18.

[Related: Meta begins automatically restricting teen users to more ‘age-appropriate’ content.]

Meta’s flurry of social media reforms come as the company continues to deal with ongoing and mounting pressure regarding its yearslong approach (or lack thereof) for protecting minors. Next week, CEO Mark Zuckerberg will be grilled—alongside the heads of X, Snap, Discord, and TikTok—at a Senate hearing on online child safety. Meanwhile, Meta faces a number of major lawsuits alleging the company ignored safety issues in favor of profiteering from young users’ data.

Knowing this, Meta isn’t done with its policy changes. In today’s update, the company also announced impending plans to implement restrictions targeting “unwanted and potentially inappropriate” images and messages from young users’ connections and friends. More information pertaining to this policy shift will purportedly arrive “later this year.”

The post Facebook and Instagram are making it harder for strangers to DM teens appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

The director of the Department of Homeland Security’s cybersecurity infrastructure protection agency confirms she was the victim of a dangerous “swatting” attempt late last month. As first reported on January 22 by cybersecurity news outlet, The Record, local police in Arlington County, VA, arrived at Cybersecurity and Infrastructure Security Agency (CISA) Director Jen Easterly’s residence around 9pm on December 30 after receiving a 911 call that falsely claimed a shooting had occurred in or near her home.

What is ‘swatting’?

“Swatting” refers to when malicious actors intentionally report nonexistent, often violent crimes at a target’s residence, with the intention of causing an aggressive, potentially harmful police response. The term originates in reference to the SWAT teams most often dispatched to handle the kinds of crimes reported by hoaxers. Although its origins reside in events such as simply calling in false bomb threats, swatting itself has grown in popularity over the years, initially through the online video gaming community. The FBI first referenced the “new phenomenon” as far back as 2008, but tactics have evolved since then to include more sophisticated methods such as hacking Ring cameras and employing “spoofing” technology to appear as though a 911 call is actually coming from a targeted residence. The technical complexity involved in Easterly’s incident is currently unclear.

[Related: Two men used Ring cameras to ‘swat’ homeowners.]

Although law enforcement officers departed Easterly’s home last month after confirming the 911 call to be a hoax, this unfortunately is not always the case. In 2017, Wichita police accidentally killed a 28-year-old after responding to false reports of a shooting and hostage situation. In that instance, the tragedy stemmed from a dispute from two online gamers with no connection to the victim after one of the players provided the other their old address.

Swatting is increasingly used to harass public and elected officials, regardless of political affiliation. The tactic’s rising popularity is considered so grave that the FBI established a national database to help track and prevent future swatting events in June 2023.

“One of the most troubling trends we have seen in recent years has been the harassment of public officials across the political spectrum, including extreme incidents involving swatting and direct personal threats,” Easterly said in a statement offered to The Record on Monday. “These incidents pose a serious risk to the individuals, their families, and in the case of swatting, to the law enforcement officers responding to the situation.”

Although Easterly described the experience as “harrowing,” she explained that swatting is now “unfortunately not unique.” CISA’s director cited bad actors recently targeting “several of our nation’s election officials” due to continued, patently false conspiracy theories and outright lies pertaining to both the 2020 election, as well as the upcoming 2024 election asserting rigged outcomes for President Biden.

In just the past few weeks, swatting attacks occurred on judges overseeing legal cases against former President Donald Trump, election officials in both Georgia and Maine, as well as both Republican and Democrat politicians. During a press conference last week, White House press secretary Karine Jean-Pierre called the trend “a danger and a risk to our society” after the White House itself faced a swatting hoax pertaining to a nonexistent fire.

CISA first formed in 2007 as a division of the Department of Homeland Security. In 2018, its responsibilities expanded to encompass national election and census cybersecurity efforts.

The post Election cybersecurity director was a victim of a ‘swatting’ attack in her home appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

AI vocal cloning technology is reportedly already muddying the waters ahead of the 2024 election. According to a statement issued by the New Hampshire attorney general’s office on Monday, a robocall campaign deployed over the weekend used an imitation of President Joe Biden’s voice to urge recipients not to vote in the state’s January 23 presidential primary.

“Voting this Tuesday only enables the Republicans in their quest to elect Donald Trump again,” an AI-generated Biden told residents over the phone. “Your vote makes a difference in November, not this Tuesday.”

The disinformation campaign’s orchestrators are currently unknown, but it comes from “obviously somebody who wants to hurt Joe Biden,” according to former New Hampshire Democratic Party chair Kathy Sullivan, speaking through NBC News’ initial exclusive report.

[Related: Deepfake audio already fools people nearly 25 percent of the time.]

A growing problem

AI-generated content including deepfaked audio, video, and imagery is a growing concern among misinformation experts. Multiple reports have warned that today’s media, internet, and social landscapes are unprepared for a likely imminent deluge of falsified “fake news” content as the 2024 presidential election intensifies. A recent study conducted by researchers at the UK’s University College London indicates AI generated audio can fool as many as 1-in-4 listeners. Over 1,600 videos uploaded to YouTube have featured deepfaked celebrities like Taylor Swift and Steve Harvey hawking “medical card” schemes and other scams, collectively amassing over 195 million views in the process. But unlike “free money” ploys from AI Oprah, the latest vocal cloning example is explicitly meant to influence the US political landscape.

“Disgraceful and an unacceptable affront to democracy”

As for this weekend’s misinfo campaign, the fake Biden falsely claimed an ongoing statewide campaign to write-in his name during New Hampshire’s primary would hurt the president’s reelection prospects. The robocall message then concluded with a phone number linked to Kathy Sullivan, resulting in a flurry of calls on Sunday evening that prompted the former New Hampshire Democratic Party Chair to report the situation to the state attorney general’s office.

“These messages appear to be an unlawful attempt to disrupt the New Hampshire Presidential Primary Election and to suppress New Hampshire voters,” the state attorney general’s office cautioned in Monday’s statement. Voters were urged to disregard the message, with the office explicitly making clear, “Voting in the New Hampshire Presidential Primary Election does not preclude a voter from additionally voting in the November General Election.”

[Related: Beware the AI celebrity clones peddling bogus ‘free money’ on YouTube.]

Representatives of former President Trump’s campaign have denied any connection to the robocall scheme. Meanwhile a spokesperson for Dean Phillips, the congressman from Minnesota challenging Biden for the Democratic Party nomination, described the scam as “wildly concerning.”

“Any effort to discourage voters is disgraceful and an unacceptable affront to democracy,” Phillips campaign representative Katie Dolan told NBC News. “The potential use of AI to manipulate voters is deeply disturbing.”

Chatbot politician stand-ins

At least some Phillips supporters, however, are embracing other AI tool tactics, despite warnings to the contrary. Late last week, The Washington Post noted that Dean.Bot—a chatbot created by a pro-Phillips SuperPAC—had been removed from OpenAI’s recently launched online store for “knowingly violating our API usage policies which disallow political campaigning.”

Despite the slap on the wrist, Phillips remains a favorite of some Silicon Valley’s top players. The SuperPAC behind Dean.Bot, We Deserve Better, was co-founded by the former chief of staff for Sam Altman, OpenAI co-founder and recently fired-and-rehired CEO. Altman himself has met with Phillips, although he has yet to endorse or formally donate to the congressman’s campaign.

An ongoing investigation

The New Hampshire AG’s statement notes a Department of Justice investigation into the AI presidential robocall is ongoing, and encourages residents to contact the Election Law Unit if they received a message beginning with AI Biden opining, “What a bunch of malarkey.”

The post A deepfake ‘Joe Biden’ robocall told voters to stay home for primary election appeared first on Popular Science.

Articles may contain affiliate links which enable us to share in the revenue of any purchases made.

This story was originally published by Grist. Sign up for Grist’s weekly newsletter here.

Imagine if you could walk from your house to anywhere you needed to go in less than 15 minutes: the pharmacy, the bakery, the gym, and then back to the bakery. In a certain, conspiracy-addled corner of the internet, this urban planning concept of “15-minute cities” gets a shady, sinister gloss. Conspiracy theorists evoke COVID restrictions and tout efforts to create walkable cities as steps toward “climate lockdowns.” They warn of a plot by the World Economic Forum to restrict people’s movements, trapping and surveilling them in their neighborhoods. 

“They want to take away your cars,” claims Clayton Morris, a former Fox News host, in a YouTube video that’s been viewed 1.7 million times.

YouTube is riddled with false claims like these, so it’s the place to document the evolution of arguments against taking action on climate change. A new report from the Center for Countering Digital Hate, a nonprofit based in London and Washington, D.C., working to stop the spread of disinformation, analyzed 12,000 videos from channels that promoted lies about climate change on YouTube over the last six years. Over that time, the reality of climate change long predicted by scientists has become increasingly difficult to dismiss. The report, released on Tuesday, found a dramatic shift from “old denial” arguments—that global warming isn’t real and isn’t caused by humans—to new arguments bent on undermining trust in climate solutions.

“The success is that the science has won this debate on anthropogenic climate change,” said Imran Ahmed, the nonprofit’s founder and CEO. “The opponents of action have shifted their attention.”

The report suggests that, rather than doing a victory lap, climate advocates may want to focus on defending climate policies and renewable energy as necessary and effective. As the world was besieged by intense heat, expansive wildfires, and catastrophic floods in recent years, YouTubers promoting disinformation increasingly embraced “new denial” narratives, such as that solar panels will destroy the economy and the environment, or that the United Nations’ Intergovernmental Panel on Climate Change is a “fraud.”

“What it is doing is creating a cohort of people who believe climate change is happening, but believe there’s no hope,” Ahmed said. People start watching YouTube at a young age—in 2020, more than half of parents in the U.S. with a child 11 years old or younger said their kid watched videos on the platform on a daily basis. New polling from the center, released alongside the study, found that a third of U.S. teens say that climate policies cause more harm than good.

Six years ago, these “new denial” claims made up 35 percent of denier’s arguments on YouTube; now, they make up 70 percent of the total. The fastest-growing assertions were that the climate movement is unreliable and that clean energy won’t work.

To get this data, the Center for Countering Digital Hate analyzed video transcripts from nearly 100 YouTube channels that spout climate denial, using an artificial intelligence tool to categorize the arguments. 

One popular source is the channel of Jordan Peterson, a Canadian psychologist and culture warrior with 7 million followers. In an interview with Alex Epstein, the author of The Moral Case for Fossil Fuels, Epstein makes the case that climate advocates can’t be trusted. “Listening to a modern environmentalist is like listening to a doctor who’s on the side of the germs, somebody who doesn’t have your best interests at heart,” Epstein says in a video entitled “The Great Climate Con” that’s been viewed a million times, reiterating a point once made in the 1990s by the economist George Reisman in an article titled “The Toxicity of Environmentalism.”

The report also points to the libertarian think tank The Heartland Institute and the media company BlazeTV, created by the former Fox News host Glenn Beck, as prominent sources of lies about climate change on YouTube. Videos from PragerU, a right-wing media outlet also known for spreading disinformation, paint solar and wind power as dangers to the environment and compare environmental activists to Nazis. Despite what the name may imply, it’s not actually a university, nor does it offer any degrees.

John Cook, a researcher at the Melbourne Centre for Behaviour Change in Australia, has documented a similar rise in attacks on climate solutions by conservative think tanks and blogs. “It’s surprising to see misinformation on YouTube shifting so quickly,” Cook said in an email. “The future of climate misinformation will be focused on attacking climate solutions, and we need to better understand those arguments and how to counter them.”

Some research has shown that climate disinformation is compelling: A recent study in Nature Human Behavior found that it’s often more persuasive to people than scientific facts. And once people latch onto a falsehood, they find it hard to let go. That’s why stopping disinformation at the source is so important, according to Ahmed. “The key right now is ensuring that we aren’t flooding our information ecosystem with nonsense and lies that make it more difficult for people to work out what’s true or not,” he said.

Together, the YouTube channels that the center focused on garnered 3.4 billion views last year. And all those views means there’s money involved: The report found that YouTube is potentially making up to $13.4 million a year in ad revenue from channels that post climate denial.

Google, which owns YouTube, promised in 2021 to ban ads on its platforms alongside content that contradicts the scientific consensus that climate change is happening and caused by humans (though it hasn’t enforced it well). To counter the latest wave of disinformation, the Center for Countering Digital Hate recommends that Google should also prohibit advertisements on content that pushes misinformation about climate solutions, so that YouTubers won’t be incentivized to publish more of it. (Content creators who partner with YouTube receive a share of the ad revenue.)

“If it wasn’t profitable, would so many people see it as being a business to produce bullshit?” Ahmed said. “We’re asking platforms to not reward liars with money and attention.”