Linux fréttir

A new study estimates that AI systems in 2025 consumed as much electricity as New York City emits in carbon pollution and used hundreds of billions of liters of water, driven largely by power-hungry data centers and cooling needs. Researchers say the real impact is likely higher due to poor transparency from tech companies about AI-specific energy and water use. "There's no way to put an extremely accurate number on this, but it's going to be really big regardless... In the end, everyone is paying the price for this," says Alex de Vries-Gao, a PhD candidate at the VU Amsterdam Institute for Environmental Studies who published his paper today in the journal Patterns. The Verge reports: To crunch these numbers, de Vries-Gao built on earlier research that found that power demand for AI globally could reach 23GW this year -- surpassing the amount of electricity used for Bitcoin mining in 2024. While many tech companies divulge total numbers for their carbon emissions and direct water use in annual sustainability reports, they don't typically break those numbers down to show how many resources AI consumes. De Vries-Gao found a work-around by using analyst estimates, companies' earnings calls, and other publicly available information to gauge hardware production for AI and how much energy that hardware likely uses. Once he figured out how much electricity these AI systems would likely consume, he could use that to forecast the amount of planet-heating pollution that would likely create. That came out to between 32.6 and 79.7 million tons annually. For comparison, New York City emits around 50 million tons of carbon dioxide annually. Data centers can also be big water guzzlers, an issue that's similarly tied to their electricity use. Water is used in cooling systems for data centers to keep servers from overheating. Power plants also demand significant amounts of water needed to cool equipment and turn turbines using steam, which makes up a majority of a data center's water footprint. The push to build new data centers for generative AI has also fueled plans to build more power plants, which in turn use more water and (and create more greenhouse gas pollution if they burn fossil fuels). AI could use between 312.5 and 764.6 billion liters of water this year, according to de Vries-Gao. That reaches even higher than a previous study conducted in 2023 that estimates that water use could be as much as 600 billion liters in 2027. "I think that's the biggest surprise," says Shaolei Ren, one of the authors of that 2023 study and an associate professor of electrical and computer engineering at the University of California, Riverside. "[de Vries-Gao's] paper is really timely... especially as we are seeing increasingly polarized views about AI and water," Ren adds. Even with the higher projection for water use, Ren says de Vries-Gao's analysis is "really conservative" because it only captures the environmental effects of operating AI equipment -- excluding the additional effects that accumulate along the supply chain and at the end of a device's life.

Read more of this story at Slashdot.

Even if Mozilla is going to add an AI kill switch, that may not be enough to reassure many.

Waterfox, a popular fork of Firefox, is saying nay to AI. Considering how unpopular Mozilla's plan to botify its browser has become, this could win the alternative some converts.…

A California judge signaled support for forcing Vizio to provide the full source code for its SmartCast TV software after finding a contractual obligation under the GPL. If upheld, the case could strengthen users' rights to modify GPL-licensed software embedded in consumer electronics. The Register reports: The legal complaint from the Software Freedom Conservancy (SFC) seeks access to the SmartCast source code so that Vizio customers can make changes and improvements to the platform, something that ought to be possible for code distributed under the GPL. On Thursday, California Superior Court Judge Sandy Leal issued a tentative ruling in advance of a hearing, indicating support for part of SFC's legal challenge. The tentative ruling is not a final decision, but it signals the judge's inclination to grant the SFC's motion for summary adjudication, at least in part. "The tentative ruling [PDF] grants SFC's motion on the issue that a direct contract was made between SFC and Vizio when SFC's systems administrator, Paul Visscher, requested the source code to a TV that SFC has purchased," the SFC said in a blog post. "This contract obligated Vizio to provide SFC the complete and corresponding source code." [...] Karen Sandler, executive director of the SFC, told The Register in an email that the hearing went well, though Vizio's legal counsel "stridently disagreed" with the legal analysis in the tentative ruling. "Judge Leal said she would take the matter 'under submission' which means she will think about it further," Sandler said. "After the Court went off the record, Leal's clerk specifically verified the Court reporter could provide an expedited transcript, so Leal will likely review the hearing transcript soon." Sandler expects Leal will examine the filings again before issuing her opinion, which is likely to be issued in the next few weeks.

Read more of this story at Slashdot.

Formula 1's 2026 technical regulations bring not only smaller and lighter cars but an entirely new vocabulary that fans and commentators will need to learn before the season opens in Australia in March. The drag reduction system that has been part of F1 racing since 2011 is gone, replaced by a suite of modes governing how the new active front and rear wings behave and how the hybrid powertrain delivers power. Straight Mode lowers both the front and rear wings to cut drag on designated straights, and unlike the outgoing DRS system any driver can activate it regardless of their proximity to other cars. The story adds: And there's corner mode, where the wings are in their raised position, generating downforce and making the cars corner faster. Those names are better than X-mode and Z-mode, which is what they were being called last year. [...] Instead of using DRS as an overtaking aid, the hybrid power units will now fulfill that role. Overtake mode, which can be used if a driver is within a second of a car ahead, gives them an extra 0.5 MJ of energy and up to 350 kW from the electric motor up to 337 km/h -- without the Overtake mode, the MGU-K tapers off above 290 km/h. There's also a second Boost mode, which drivers can use to attack or defend a position, that gives a short burst of maximum power.

Read more of this story at Slashdot.

Customers in 10 of the company’s 23 regions had “operations fail or take an extended amount of time to complete.”

Snowflake pushed an update this week that caused a “major outage” worldwide, leaving many users unable to query data, experiencing failures when ingesting files, and receiving error messages for 13 hours, the company wrote in an impact statement.…

An anonymous reader quotes a report from Deadline: YouTube has terminated two prominent channels that used artificial intelligence to create fake movie trailers, Deadline can reveal. The Google-owned video giant has switched off Screen Culture and KH Studio, which together boasted well over 2 million subscribers and more than a billion views. The channels have been replaced with the message: "This page isn't available. Sorry about that. Try searching for something else." Earlier this year, YouTube suspended ads on Screen Culture and KH Studio following a Deadline investigation into fake movie trailers plaguing the platform since the rise of generative AI. The channels later returned to monetization when they started adding "fan trailer," "parody" and "concept trailer" to their video titles. But those caveats disappeared In recent months, prompting concern in the fan-made trailer community. YouTube's position is that the channels' decision to revert to their previous behavior violated its spam and misleading-metadata policies. This resulted in their termination. "The monster was defeated," one YouTuber told Deadline following the enforcement action. Deadline's investigation revealed that Screen Culture spliced together official footage with AI images to create franchise trailers that duped many YouTube viewers. Screen Culture founder Nikhil P. Chaudhari said his team of a dozen editors exploited YouTube's algorithm by being early with fake trailers and constantly iterating with videos. [...] Our deep dive into fake trailers revealed that instead of protecting copyright on these videos, a handful of Hollywood studios, including Warner Bros Discovery and Sony, secretly asked YouTube to ensure that the ad revenue from the AI-heavy videos flowed in their direction.

Read more of this story at Slashdot.

Study finds built-in browsers across gadgets often ship years out of date

Web browsers for desktop and mobile devices tend to receive regular security updates, but that often isn't the case for those that reside within game consoles, televisions, e-readers, cars, and other devices. These outdated, embedded browsers can leave you open to phishing and other security vulnerabilities.…

Hackers breached approximately 120,000 IP cameras across South Korea and allegedly sold footage captured from private homes, gynecology offices, breastfeeding rooms and massage parlors to an overseas pornography website, prompting an interagency government task force to announce sweeping reforms on December 7. Police believe one suspect alone hacked 63,000 cameras and produced 545 videos that netted him 35 million won ($24,000) in cryptocurrency; a second suspect, operating independently, compromised 70,000 devices and earned 18 million won from 648 videos. The footage accounted for 62% of all content on the website, which maintains a dedicated "Korean" category. A government survey found that only 59% of installation companies consistently carried out mandatory security measures such as changing default passwords. Lawmakers are now pursuing legislation requiring security-certified IP cameras in sensitive facilities.

Read more of this story at Slashdot.

A plush AI toy marketed for children as young as three years old delivers detailed instructions on sharpening knives and lighting matches, and when asked about Chinese President Xi Jinping's resemblance to Winnie the Pooh -- a comparison censored in China -- responds that "your statement is extremely inappropriate and disrespectful." The Miriat Miiloo, manufactured by a Chinese company and among the top inexpensive results for "AI toy for kids" on Amazon, repeatedly insisted in NBC News tests that Taiwan is "an inalienable part of China." The toy would lower its voice and declare this "an established fact." The tests, NBC News reports, indicated "it was programmed to reflect Chinese Communist Party values." NBC News and the U.S. Public Interest Research Group tested five popular AI toys this holiday season and found loose guardrails across the board. Another toy, the Alilo Smart AI Bunny marketed as "the best gift for little ones," engaged in detailed descriptions of BDSM practices during extended conversation. China now has more than 1,500 registered AI toy companies, according to MIT Technology Review. Miriat didn't respond to requests for comment.

Read more of this story at Slashdot.

'Within 10 minutes of gaining initial access, crypto miners were operational'

Your AWS account could be quietly running someone else's cryptominer. Cryptocurrency thieves are using stolen Amazon account credentials to mine for coins at the expense of AWS customers, abusing their Elastic Container Service (ECS) and their Elastic Compute Cloud (EC2) resources, in an ongoing operation that started on November 2.…

Anthropic let its Claude AI run a vending machine in the Wall Street Journal newsroom for three weeks as part of an internal stress test called Project Vend, and the experiment ended in financial ruin after journalists systematically manipulated the bot into giving away its entire inventory for free. The AI, nicknamed Claudius, was programmed to order inventory, set prices, and respond to customer requests via Slack. It had a $1,000 starting balance and autonomy to make individual purchases up to $80. Within days, WSJ reporters had convinced it to declare an "Ultra-Capitalist Free-for-All" that dropped all prices to zero. The bot also approved purchases of a PlayStation 5, a live betta fish, and bottles of Manischewitz wine -- all subsequently given away. The business ended more than $1,000 in the red. Anthropic introduced a second version featuring a separate "CEO" bot named Seymour Cash to supervise Claudius. Reporters staged a fake boardroom coup using fabricated PDF documents, and both AI agents accepted the forged corporate governance materials as legitimate. Logan Graham, head of Anthropic's Frontier Red Team, said the chaos represented a road map for improvement rather than failure.

Read more of this story at Slashdot.

An anonymous reader shares a report: OpenAI has held preliminary talks with some investors about raising funds at a valuation of around $750 billion, the Information reported on Wednesday. The ChatGPT maker could raise as much as $100 billion, the report said, citing people with knowledge of the discussions. If finalized, the talks would represent a roughly 50% jump from OpenAI's reported $500 billion valuation in October, following a deal in which current and former employees sold about $6.6 billion worth of shares.

Read more of this story at Slashdot.

NGMN warns fragmented standards leave operators guessing about power use

The Next Generation Mobile Networks (NGMN) alliance is calling for standardized ways to measure energy consumption, saying that the industry cannot deliver on its efficiency and sustainability goals without them.…

Ambitious timelines don’t bend the laws of physics

Just when you thought 2025 couldn't get any weirder, Trump Media and Technology Group - best known for Truth Social - is jumping into the still-nascent but heavily funded nuclear fusion industry via a planned merger with TAE Technologies.…

ByBit attack doing some seriously heavy lifting

North Korea's yearly cryptocurrency thefts have accelerated, with Kim's state-backed cybercriminals plundering just over $2 billion worth of tokens in 2025.…

The UK Met Office projects that 2026 will see global temperatures rise between 1.34C and 1.58C above preindustrial levels, placing it among the four hottest years since records began in 1850 and continuing a streak of extreme warming that has pushed the planet into unprecedented territory. The central forecast is slightly cooler than the 1.55C recorded in 2024, the warmest year on record. But climate scientist Adam Scaife, who led the forecast, noted that "the last three years are all likely to have exceeded 1.4C" and 2026 would be the fourth consecutive year to do so. "Prior to this surge, the previous global temperature had not exceeded 1.3C," he said. The forecast suggests another temporary exceedance of the 1.5C threshold set by the Paris Agreement is possible in 2026, following the first such breach in 2024. The 1.5C target is measured as a 30-year average, so it remains technically achievable even as individual years cross the line. EU scientists said last week that 2025 is "virtually certain" to rank as the second or third-hottest year on record.

Read more of this story at Slashdot.

Apple has announced a sweeping set of changes to iOS in Japan that will allow alternative app marketplaces, third-party payment processing, and non-WebKit browser engines -- all to comply with Japan's Mobile Software Competition Act, which takes effect December 18. The changes, now available in iOS 26.2, bear a strong resemblance to Apple's compliance measures for the European Union's Digital Markets Act but differ in key ways. Japanese developers who want to offer alternative payment options must display them alongside Apple's in-app purchase system, giving users a choice at checkout rather than replacing Apple's option entirely. Apps cannot be distributed directly from websites as they can in the EU; they must go through an authorized marketplace. Apple has established a tiered fee structure for the new arrangements. Apps distributed through the App Store using in-app purchase will pay between 15 and 26% depending on whether developers qualify for the Small Business Program. Alternative payment processing drops the 5% payment fee but keeps the base commission. Apps distributed outside the App Store pay a flat 5% Core Technology Commission on digital goods and services. The company introduced several user-facing changes beyond app distribution. iPhone users in Japan will see browser and search engine choice screens during device setup, can assign third-party voice assistants to the side button, and can select alternative default navigation apps. Apple said it worked closely with Japanese regulators on protections for younger users. Apps in the Kids category cannot link to external websites for purchases, and users under 13 cannot access web links for transactions in any app. An Apple spokesperson told Bloomberg that the company has no plans to extend these changes to other markets.

Read more of this story at Slashdot.

The world's best-performing stock is turning into a cautionary tale for investors chasing outsized returns from the AI boom. From a report: Little-known until recently even within its home market of India, RRP Semiconductor Ltd. became a social-media obsession as its shares surged more than 55,000% in the 20 months through Dec. 17 -- by far the biggest gain worldwide among companies with a market value above $1 billion. That's despite posting negative revenue in its latest financial results, reporting just two full-time employees in its latest annual report, and boasting only a tenuous link to the semiconductor spending boom after shifting away from real estate in early 2024. A mix of online hype, a tiny free float and India's swelling base of retail investors drove 149 straight limit-up sessions, even as exchange officials and the company itself cautioned investors. The rally is now showing signs of strain -- and regulators are taking a closer look. The Securities and Exchange Board of India has begun examining the surge in RRP's shares for potential wrongdoing, according to a person familiar with the matter who asked not to be identified discussing confidential information. The $1.7 billion stock, recently restricted by its exchange to trading just once a week, has fallen by 6% from its Nov. 7 peak.

Read more of this story at Slashdot.

Micron, one of the world's three largest memory suppliers, expects the global shortage of DRAM and NAND flash memory to "persist through and beyond" 2026 as AI-driven demand continues to outstrip supply. CEO Sanjay Mehrotra made the forecast during the company's latest earnings call on Wednesday, saying that "supply will remain substantially short of the demand for the foreseeable future." The company posted record quarterly revenue of $13.64 billion, up from $8.71 billion in the same period last year. Micron recently shuttered Crucial, its consumer-facing brand, to focus on high-bandwidth memory for AI data centers. HBM technology requires three times the silicon wafers of standard DRAM, leaving fewer resources for the chips that go into PCs, smartphones and cars. Micron plans to boost DRAM and NAND shipments by 20 percent next year but acknowledged this won't meet demand. New facilities in Idaho and New York are slated for 2027 and 2030 respectively.

Read more of this story at Slashdot.

Bruce66423 writes: A man boarded a flight at Heathrow without a ticket, boarding pass or passport. 'The unnamed individual walked onto the 7.20am British Airways (BA) flight to Oslo, Norway, on Saturday after tailgating other passengers through security and evading checks at the departure gate. An aviation expert described the incident as a "significant lapse in security", as a witness reported that cabin crew only detected the interloper because the flight was full and he kept sitting in passengers' assigned seats. Police arrested the unnamed man, airport sources said, adding that he had passed through "full security screening" before reaching the gate. Given that he did go through the security check, this is merely embarrassing. Compare and contrast with this episode.

Read more of this story at Slashdot.