Scientists are putting Einstein’s claim that the speed of light is constant to the test. While researchers found no evidence that light’s speed changes with energy, this null result dramatically tightens the constraints on quantum-gravity theories that predict even the tiniest violations. ScienceDaily reports: Special relativity rests on the principle that the laws of physics remain the same for all observers, regardless of how they are moving relative to one another. This idea is known as Lorentz invariance. Over time, Lorentz invariance became a foundational assumption in modern physics, especially within quantum theory. […] One prediction shared by several Lorentz-invariance-violating quantum gravity models is that the speed of light may depend slightly on a photon’s energy. Any such effect would have to be tiny to match existing experimental limits. However, it could become detectable at the highest photon energies, specifically in very-high-energy gamma rays.

A research team led by former UAB student Merce Guerrero and current IEEC PhD student at the UAB Anna Campoy-Ordaz set out to test this idea using astrophysical observations. The team also included Robertus Potting from the University of Algarve and Markus Gaug, a lecturer in the Department of Physics at the UAB who is also affiliated with the IEEC. Their approach relies on the vast distances light travels across the universe. If photons of different energies are emitted at the same time from a distant source, even minuscule differences in their speeds could build up into measurable delays by the time they reach Earth.

Using a new statistical technique, the researchers combined existing measurements of very-high-energy gamma rays to examine several Lorentz-invariance-violating parameters favored by theorists within the Standard Model Extension (SME). The goal was ambitious. They hoped to find evidence that Einstein’s assumptions might break down under extreme conditions. Once again, Einstein’s predictions held firm. The study did not detect any violation of Lorentz invariance. Even so, the results are significant. The new analysis improves previous limits by an order of magnitude, sharply narrowing where new physics could be hiding.

Read more of this story at Slashdot.

Meta has signed long-term nuclear power deals totaling more than 6 gigawatts to fuel its data centers: “one from a startup, one from a smaller energy company, and one from a larger company that already operates several nuclear reactors in the U.S,” reports TechCrunch. From the report: Oklo and TerraPower, two companies developing small modular reactors (SMR), each signed agreements with Meta to build multiple reactors, while Vistra is selling capacity from its existing power plants. […] The deals are the result of a request for proposals that Meta issued in December 2024, in which Meta sought partners that could add between 1 to 4 gigawatts of generating capacity by the early 2030s. Much of the new power will flow through the PJM interconnection, a grid which covers 13 Mid-Atlantic and Midwestern states and has become saturated with data centers.

The 20-year agreement with Vistra will have the most immediate impact on Meta’s energy needs. The tech company will buy a total of 2.1 gigawatts from two existing nuclear power plants, Perry and Davis-Besse in Ohio. As part of the deal, Vistra will also add capacity to those power plants and to its Beaver Valley power plant in Pennsylvania. Together, the upgrades will generate an additional 433 MW and are scheduled to come online in the early 2030s.

Meta is also buying 1.2 gigawatts from young provider Oklo. Under its deal with Meta, Oklo is hoping to start supplying power to the grid as early as 2030. The SMR company went public via SPAC in 2023, and while Oklo has landed a large deal with data center operator Switch, it has struggled to get its reactor design approved by the Nuclear Regulatory Commission. If Oklo can deliver on its timeline, the new reactors would be built in Pike County, Ohio. The startup’s Aurora Powerhouse reactors each produce 75 megawatts of electricity, and it will need to build more than a dozen to fulfill Meta’s order. TerraPower is a startup co-founded by Bill Gates, and it is aiming to start sending electricity to Meta as early as 2032.

Read more of this story at Slashdot.

An anonymous reader quotes a report from Wired: [P]erhaps AI can, in fact, learn in a more human way — by figuring out interesting questions to ask itself and attempting to find the right answer. A project from Tsinghua University, the Beijing Institute for General Artificial Intelligence (BIGAI), and Pennsylvania State University shows that AI can learn to reason in this way by playing with computer code. The researchers devised a system called Absolute Zero Reasoner (AZR) that first uses a large language model to generate challenging but solvable Python coding problems. It then uses the same model to solve those problems before checking its work by trying to run the code. And finally, the AZR system uses successes and failures as a signal to refine the original model, augmenting its ability to both pose better problems and solve them.

The team found that their approach significantly improved the coding and reasoning skills of both 7 billion and 14 billion parameter versions of the open source language model Qwen. Impressively, the model even outperformed some models that had received human-curated data. […] A key challenge is that for now the system only works on problems that can easily be checked, like those that involve math or coding. As the project progresses, it might be possible to use it on agentic AI tasks like browsing the web or doing office chores. This might involve having the AI model try to judge whether an agent’s actions are correct. One fascinating possibility of an approach like Absolute Zero is that it could, in theory, allow models to go beyond human teaching. “Once we have that it’s kind of a way to reach superintelligence,” [said Zilong Zheng, a researcher at BIGAI who worked on the project].

Read more of this story at Slashdot.

Experts interviewed by NBC News warn that the rapid spread of AI-generated images and videos is accelerating an online trust breakdown, especially during fast-moving news events where context is scarce. From the report: President Donald Trump’s Venezuela operation almost immediately spurred the spread of AI-generated images, old videos and altered photos across social media. On Wednesday, after an Immigration and Customs Enforcement officer fatally shot a woman in her car, many online circulated a fake, most likely AI-edited image of the scene that appears to be based on real video. Others used AI in attempts to digitally remove the mask of the ICE officer who shot her.

The confusion around AI content comes as many social media platforms, which pay creators for engagement, have given users incentives to recycle old photos and videos to ramp up emotion around viral news moments. The amalgam of misinformation, experts say, is creating a heightened erosion of trust online — especially when it mixes with authentic evidence. “As we start to worry about AI, it will likely, at least in the short term, undermine our trust default — that is, that we believe communication until we have some reason to disbelieve,” said Jeff Hancock, founding director of the Stanford Social Media Lab. “That’s going to be the big challenge, is that for a while people are really going to not trust things they see in digital spaces.”

Though AI is the latest technology to spark concern about surging misinformation, similar trust breakdowns have cycled through history, from election misinformation in 2016 to the mass production of propaganda after the printing press was invented in the 1400s. Before AI, there was Photoshop, and before Photoshop, there were analog image manipulation techniques. Fast-moving news events are where manipulated media have the biggest effect, because they fill in for the broad lack of information, Hancock said. “In terms of just looking at an image or a video, it will essentially become impossible to detect if it’s fake. I think that we’re getting close to that point, if we’re not already there,” said Hancock. “The old sort of AI literacy ideas of ‘let’s just look at the number of fingers’ and things like that are likely to go away.”

Renee Hobbs, a professor of communication studies at the University of Rhode Island, added: “If constant doubt and anxiety about what to trust is the norm, then actually, disengagement is a logical response. It’s a coping mechanism. And then when people stop caring about whether something’s true or not, then the danger is not just deception, but actually it’s worse than that. It’s the whole collapse of even being motivated to seek truth.”

Read more of this story at Slashdot.

Intel CEO Lip-Bu Tan says the company is “going big time” into its 14A (1.4nm-class) process, signaling confidence in yields and hinting at at least one external foundry customer. Tom’s Hardware reports: Intel’s 14A is expected to be production-ready in 2027, with early versions of process design kit (PDK) coming to external customers early this year. To that end, it is good to hear Intel’s upbeat comments about 14A. Also, Tan’s phrasing ‘the customer’ could indicate that Intel has at least one external client for 14A, implying that Intel Foundry will produce 14A chips for Intel Products and at least one more buyer.

The 14A production node will introduce Intel’s 2nd Generation RibbonFET GAA transistors; 2nd Gen BSPDN called PowerDirect that will connect power directly to source and drain of transistors, enabling better power delivery (e.g., reducing transient voltage droop or clock stretching) and refined power controls; and Turbo Cells that optimize critical timing paths using high-drive, double-height cells within dense standard cell libraries, which boost speed without major area or power compromises.

Yet, there is another aspect of Intel’s 14A manufacturing process that is particularly important for the chipmaker: its usage by external customers. With 18A, the company has not managed to land a single major external client that demands decent volumes. While 18A will be used by Intel itself as well as by Microsoft and the U.S. Department of Defense, only Intel will consume significant volumes. For 14A, Intel hopes to land at least one more external customer with substantial volume requirements, as this will ensure that Intel will recoup its investments in the development of such an advanced node.

Read more of this story at Slashdot.

SpaceX today received US permission to launch another 7,500 second-generation Starlink satellites, bringing its total authorization to 15,000 Gen2 satellites including those previously approved.

“Under this grant, SpaceX is authorized to construct, deploy, and operate an additional 7,500 Gen2 Starlink satellites, bringing the total to 15,000 satellites worldwide,” the Federal Communications Commission announced today. “This expansion will enable SpaceX to deliver high-speed, low-latency Internet service globally, including enhanced mobile and supplemental coverage from space.”

The FCC gave SpaceX permission for the first set of 7,500 satellites in December 2022. The agency deferred action on the rest of the second-generation constellation at the time and limited the first batch to certain altitudes, saying it needed to “address concerns about orbital debris and space safety” before approving the full bunch.

Read full article

Comments

Microsoft is testing a new Windows policy that lets IT administrators uninstall Microsoft Copilot from managed devices. The change rolls out via Windows Insider builds and works through standard management tools like Intune and SCCM. BleepingComputer reports: The new policy will apply to devices where the Microsoft 365 Copilot and Microsoft Copilot are both installed, the Microsoft Copilot app was not installed by the user, and the Microsoft Copilot app was not launched in the last 28 days. “Admins can now uninstall Microsoft Copilot for a user in a targeted way by enabling a new policy titled RemoveMicrosoftCopilotApp,” the Windows Insider team said.

“If this policy is enabled, the Microsoft Copilot app will be uninstalled, once. Users can still re-install if they choose to. This policy is available on Enterprise, Pro, and EDU SKUs. To enable this policy, open the Group policy editor and go to: User Configuration -> Administrative Templates -> Windows AI -> Remove Microsoft Copilot App.”

Read more of this story at Slashdot.

It took a while, but a consensus has emerged in Europe that the continent’s space industry needs to develop reusable rockets. How to do it and how much to spend on it remain unresolved questions.

Much of the discourse around reusable rockets in Europe has focused on developing a brand new rocket that might eventually replace the Ariane 6, which debuted less than two years ago but still uses the use it and lose it model embraced by the launch industry for most of the Space Age.

The European Space Agency (ESA) is offering money to emerging rocket companies in Europe to prove their small satellite launchers can do the job. ESA is also making money available to incentivize rocket upgrades to haul heavier cargo into orbit. ESA, the European Commission, and national governments are funding rocket hoppers to demonstrate vertical takeoff and vertical landing technologies. While there is significant money behind these efforts, the projects are not unified, and progress has been slow.

Read full article

Comments

Longtime Slashdot reader chicksdaddy writes: CES, the Consumer Electronics Show, isn’t just about shiny new gadgets. As AP reports, this year brought back the fifth annual Worst in Show anti-awards, calling out the most harmful, wasteful, invasive, and unfixable tech at the Las Vegas show. The coalition behind the awards — including Repair.org, iFixit, EFF, PIRG, Secure Repairs, and others — put the spotlight on products that miss the point of innovation and make life worse for users.

2026 Worst in Show winners include:

Overall (and Repairability): Samsung’s AI-packed Family Hub Fridge — over-engineered, hard to fix, and trying to do everything but keep food cold.

Privacy: Amazon Ring AI — expanding surveillance with features like facial recognition and mobile towers.

Security: Merach UltraTread treadmill — an AI fitness coach that also hoovers up sensitive data with weak security guarantees, including a privacy policy that declares the company “cannot guarantee the security of your personal information” (!!).

Environmental Impact: Lollipop Star — a single-use, music-playing electronic lollipop that epitomizes needless e-waste.

Enshittification: Bosch eBike Flow App — pushing lock-in and digital restrictions that make gear worse over time.

“Who Asked For This?”: Bosch Personal AI Barista — a voice-assistant coffee maker that nobody really wanted.

People’s Choice: Lepro Ami AI Companion — an overhyped “soulmate” cam that creeps more than it comforts.

The message? Not all tech is progress. Some products add needless complexity, threaten privacy, or throw sustainability out the window — and the industry’s watchdogs are calling them out.

Read more of this story at Slashdot.

A measles outbreak in South Carolina that began in October continues to rage, with the state health department reporting Friday that nearly 100 new cases have been identified just in the last three days.

In a regularly scheduled update this afternoon, the health department said 99 cases were identified since Tuesday, bringing the outbreak total to 310 cases. There are currently 200 people in quarantine and nine in isolation. However, the outbreak is expanding so quickly and with so many exposure sites that health officials are struggling to trace cases and identify people at risk.

“An increasing number of public exposure sites are being identified with likely hundreds more people exposed who are not aware they should be in quarantine if they are not immune to measles,” Dr. Linda Bell, state epidemiologist and the health department’s incident commander for the measles outbreak, said in the announcement. “Previous measles transmission studies have shown that one measles case can result in up to 20 new infections among unvaccinated contacts.”

Read full article

Comments

If you’ve ever been intrigued by the mystery of a dusty cassette you found in a thrift shop—or if you’re just looking for a new time-sink—you have to check out Intertapes, a website that digitizes “found cassettes” sent in by users all over the world, then posts them in full for anyone to listen to.

The catalog is small at the moment—only 14 cassettes—but already really interesting. There’s a bootleg cassette of music played at a Spanish nightclub in the late 1990s (lots of squelchy noises and relentless bass) and a 90-minute recording of New York hip hop station WBLS captured in ’94 (Warren G.’s “Regulate” represent), amid more mysterious choices, like this haunting recording from a “destroyed cassette tape found on the side of the coast highway near Heraklion” in Greece; this tape full of ominous noises found in a parking lot in Tbilisi, Georgia; tape of binary code from Barcelona; and a cassette recorded in the USSR featurng 1970s pop hits.

I love how each cassette is treated like an important archeological object, because in a way, they are—discreet time capsules made more poignant by the hiss and warp that speaks to the time that’s passed since this audio was captured and the ephemeral nature of analogue recording. From musical snapshots to accidental field recordings, these tapes are fascinating for there mere existence in the modern day, where the question of who recorded them and why adds a layer of mystery to each one.

The ongoing cassette tape revival

Intertapes could be viewed as a reflection of the growing cassette tape revival, a movement that celebrates the outdated format. Since they hit the market in 1963, audiophiles have generally considered cassette tapes an inferior format to vinyl—tapes are more rugged than records, but the sound quality is markedly worse. The spread of CDs and streaming music pretty much killed off commercial cassette releases by the early 2000s, and it’s easy to see why: Digital music doesn’t hiss or degrade. Cassettes have a more narrow dynamic range. You can instantly select tracks on a CD or MP3 player, and it will never play at a slightly wrong speed, unspool, or melt on your car dashboard. Bonus: You never have to rewind them to hear a song again.

Most people didn’t see it at the time, but when tapes slipped into obsolescence, we lost something real and tangible. Dropouts, distortion, and warp are evidence of life. Cassette tape compression is a unique sonic aesthetic that conveys warmth and nostalgia. And then there’s the way they impart meaning into the act of “listening to music.” Starting a Spotify stream is frictionless, optimized, and weightless, while cassettes are physical objects with histories that defy the disconnection of the digital space. You own the music on tapes in a way you never own information being served to you by a tech company. A friend handing you a cassette of their favorite songs is meaningful in a way a link to playlist will never be, and your Spotify playlist will never be found by the side of the road near Heraklion, to be pondered over by future people.

Yes, by digitizing them, Intertapes is removing some of the qualities that make these recording special—but it’s also preserving them, at least for now (if you’ve ever tried following a decades-old weblink, you know the internet is ephemeral too).

How to submit your own tapes to Intertape

If you’re of a certain age, you probably have a dusty cassette or two hanging around somewhere. Don’t let it molder in a desk drawer. Describe the origin of your recording and a background story, scan a picture of your tape, and email connect@intertapes.net to arrange you submission to the site. This collection really deserves to grow.