Mike Swanson, commenting on modern software's intrusive, attention-seeking behavior: What if your car worked like so many apps? You're driving somewhere important...maybe running a little bit late. A few minutes into the drive, your car pulls over to the side of the road and asks: "How are you enjoying your drive so far?" Annoyed by the interruption, and even more behind schedule, you dismiss the prompt and merge back into traffic. A minute later it does it again. "Did you know I have a new feature? Tap here to learn more." It blocks your speedometer with an overlay tutorial about the turn signal. It highlights the wiper controls and refuses to go away until you demonstrate mastery. Ridiculous, of course. And yet, this is how a lot of modern software behaves. Not because it's broken, but because we've normalized an interruption model that would be unacceptable almost anywhere else.

Read more of this story at Slashdot.

An anonymous reader quotes a report from Variety: In the future, studios that use synthetic actors in place of humans might have to pay a royalty into a union fund. That's one of the ideas kicking around as SAG-AFTRA prepares to sit down with the studios on Feb. 9. Artificial intelligence was central to the 2023 actors strike, and it's only gotten more urgent since. Social media is awash in slop, while user-made videos of Leia and Elsa are soon to debut on Disney+. And then there's Tilly Norwood -- the digital creation that crystallized AI fears last fall. Though SAG-AFTRA won some AI protections in the strike, it can't stop Tilly and her ilk from taking actors' jobs. As negotiations with studios begin early ahead of the June contract deadline, AI remains the most existential concern. Actors are also pushing to revisit streaming residuals, arguing that current "success bonuses" fall far short of the rerun-based income that once sustained middle-class careers. They also note the strain caused from long streaming hiatuses, exclusivity clauses, and self-taped auditions.

Read more of this story at Slashdot.

Longtime Slashdot reader schwit1 shares a report from CBS News: A former Google engineer has been found guilty on multiple federal charges for stealing the tech giant's trade secrets on artificial intelligence to benefit Chinese companies he secretly worked for, federal prosecutors said. According to the U.S. Attorney's Office for the Northern District of California, a jury on Thursday convicted Linwei Ding on seven counts of economic espionage and seven counts of theft of trade secrets, following an 11-day trial. The 38-year-old, also known as Leon Ding, was hired by Google in 2019 and was a resident of Newark. According to evidence presented at trial, Ding stole more than 2,000 pages of confidential information containing Google AI trade secrets between May 2022 and April 2023. He uploaded the information to his personal Google Cloud account. Around the same time, Ding secretly affiliated himself with two Chinese-based technology companies. Around June 2022, prosecutors said Ding was in discussions to be the chief technology officer for an early-stage tech company. Several months later, he was in the process of founding his own AI and machine learning company in China, acting as the company's CEO. Prosecutors said Ding told investors that he could build an AI supercomputer by copying and modifying Google's technology. In late 2023, prosecutors said Ding downloaded the trade secrets to his own personal computer before resigning from Google. According to the superseding indictment, Google uncovered the uploads after finding out that Ding presented himself as CEO of one of the companies during an Beijing investor conference. Around the same time, Ding told his manager he was leaving the company and booked a one-way flight to Beijing. "Silicon Valley is at the forefront of artificial intelligence innovation, pioneering transformative work that drives economic growth and strengthens our national security. The jury delivered a clear message today that the theft of this valuable technology will not go unpunished," U.S. Attorney Craig Missakian said in a statement.

Read more of this story at Slashdot.

A large Swedish study of 100,000 women found that using AI to assist radiologists reading mammograms reduced the rate of aggressive "interval" breast cancers by 12%. CBC News reports: For the study -- published in Thursday's issue of the medical journal The Lancet -- more than 100,000 women had mammography screenings. Half were supported by AI and the rest had their mammograms reviewed by two different radiologists, a standard practice in much of Europe known as double reading. It is not typically used in Canada, where usually one radiologist checks mammograms. The study looked at the rates of interval cancer, the term doctors use for invasive tumors that appear between routine mammograms. They can be harder to detect and studies have shown that they are more likely to be aggressive with a poorer prognosis. The rate of interval cancers decreased by 12 percent in the groups where the AI screening was implemented, the study showed. [...] Throughout the two-year study, the mammograms that were supported by AI were triaged into two different groups. Those that were determined to be low risk needed only one radiologist to examine them, while those that were considered high risk required two. The researchers reported that numerically, the AI-supported screening resulted in 11 fewer interval cancers than standard screening (82 versus 93, or 12 per cent). "This is really a way to improve an overall screening test," [said lead author, Dr. Kristina Lang]. She acknowledged that while the study found a decrease in interval cancer, longer-term studies are needed to find out how AI-supported screening might impact mortality rates. The screenings for the study all took place at one centre in Sweden, which the researchers acknowledged is a limitation. Another is that the race and ethnicity of the participants were not recorded. The next step, Lang said, will be for Swedish researchers to determine cost-effectiveness.

Read more of this story at Slashdot.

An anonymous reader quotes a report from the Guardian: The UK could introduce a universal basic income (UBI) to protect workers in industries that are being disrupted by AI, the investment minister Jason Stockwood has said. "Bumpy" changes to society caused by the introduction of the technology would mean there would have to be "some sort of concessionary arrangement with jobs that go immediately", Lord Stockwood said. The Labour peer told the Financial Times: "Undoubtedly we're going to have to think really carefully about how we soft-land those industries that go away, so some sort of [universal basic income], some sort of lifelong mechanism as well so people can retrain." A universal basic income is not part of official government policy, but when asked whether people in government were considering the need for UBI, Stockwood told the FT: "People are definitely talking about it." [...] While he has previously been a vocal proponent of a wealth tax in the UK, Stockwood told the FT he had not repeated his calls for the government to go further on taxing the rich. However, he added: "If you make your money and the first thing you do is you speak to a tax adviser to ask: 'Where can we pay the lowest tax?' we don't want those people in this country, I'd suggest, because you're not committed to your communities and the long-term success in this country."

Read more of this story at Slashdot.

Comcast's attempt to slow broadband customer losses still isn't stopping the bleeding as fiber and fixed wireless competition intensifies. In Q4 2025 alone, Comcast lost 181,000 broadband subscribers, even as it leans harder into wireless bundling and other business lines like Peacock and theme parks. Ars Technica reports: The Q4 net loss is more than the 176,000 loss predicted by analysts, although not as bad as the 199,000-customer loss that spurred [Comcast President Mike Cavanagh's] comment about Comcast "not winning in the marketplace" nine months ago. The Q4 2025 loss reported today is also worse than the 139,000-customer loss in Q4 2024 and the 34,000-customer loss in Q4 2023. "Subscriber losses were 181,000, as the early traction we are seeing from our new initiatives was more than offset by continued competitive intensity," Comcast CFO Jason Armstrong said during an earnings call today, according to a Motley Fool transcript. Comcast's residential broadband customers dropped to 28.72 million, while business broadband customers dropped to 2.54 million, for a total of 31.26 million. Armstrong said that average revenue per user grew 1.1 percent, "consistent with the deceleration that we had previewed reflecting our new go-to-market pricing, including lower everyday pricing and strong adoption of free wireless lines." Armstrong expects average revenue per user to continue growing slowly "for the next couple of quarters, driven by the absence of a rate increase, the impact from free wireless lines, and the ongoing migration of our base to simplified pricing." Comcast Connectivity & Platforms chief Steve Croney said the firm is facing "a more competitive environment from fiber" and continued competition from fixed wireless. "The market is going to remain intensely competitive," he said.

Read more of this story at Slashdot.

Longtime Slashdot reader devnulljapan writes: In 2012, Canada passed anti-circumvention law Bill C-11, cut-and-pasted from the U.S. DMCA, in return for access to U.S. markets without tariffs. Trump has tariffed Canada anyway, so Cory Doctorow suggests it sounds like like a good idea to ditch Bill C-11 and turn Canada into a "Disenshittification Nation" and go into the business of "disenshittify[ing] America's defective tech exports." Some of the specific ways Canada could respond include legalize jailbreaking, allow alternative app stores/clients, force companies to offer repair tools, and open firmware that break monopoly lock-ins. Cory's pitch is equal parts economic strategy (capture the rents Big Tech extracts) and national security (reduce dependence on U.S. tech stacks that can be switched off or weaponized).

Read more of this story at Slashdot.

A group of Linux gaming-focused distros and developers have formed the Open Gaming Collective to pool work on shared components like kernels, input systems, and Valve tooling. The Verge reports: Universal Blue, developer of the gaming-focused Linux distribution Bazzite, announced on Wednesday that its helping to form the OGC with several other groups, which will collaborate on improvements to the Linux gaming ecosystem and âoecentralize efforts around critical components like kernel patches, input tooling, and essential gaming packages such as gamescope." The other founding members of the OGC include Nobara, ChimeraOS, Playtron, Fyra Labs, PikaOS, ShadowBlip, and Asus Linux. [...] It's worth noting that this will mean some changes to Bazzite, which is switching to the OGC kernel, replacing HHD with InputPlumber as its input framework, and integrating features like RGB and fan control into the Steam UI. Bazzite also added that, "We'll be sharing patches we've made to various Valve packages with the OGC and attempting to upstream everything we can."

Read more of this story at Slashdot.

An anonymous reader quotes a report from Wired: Earlier this month, Joseph Thacker's neighbor mentioned to him that she'd preordered a couple of stuffed dinosaur toys for her children. She'd chosen the toys, called Bondus, because they offered an AI chat feature that lets children talk to the toy like a kind of machine-learning-enabled imaginary friend. But she knew Thacker, a security researcher, had done work on AI risks for kids, and she was curious about his thoughts. So Thacker looked into it. With just a few minutes of work, he and a web security researcher friend named Joel Margolis made a startling discovery: Bondu's web-based portal, intended to allow parents to check on their children's conversations and for Bondu's staff to monitor the products' use and performance, also let anyone with a Gmail account access transcripts of virtually every conversation Bondu's child users have ever had with the toy. Without carrying out any actual hacking, simply by logging in with an arbitrary Google account, the two researchers immediately found themselves looking at children's private conversations, the pet names kids had given their Bondu, the likes and dislikes of the toys' toddler owners, their favorite snacks and dance moves. In total, Margolis and Thacker discovered that the data Bondu left unprotected -- accessible to anyone who logged in to the company's public-facing web console with their Google username -- included children's names, birth dates, family member names, "objectives" for the child chosen by a parent, and most disturbingly, detailed summaries and transcripts of every previous chat between the child and their Bondu, a toy practically designed to elicit intimate one-on-one conversation. More than 50,000 chat transcripts were accessible through the exposed web portal. When the researchers alerted Bondu about the findings, the company acted to take down the console within minutes and relaunched it the next day with proper authentication measures. "We take user privacy seriously and are committed to protecting user data," Bondu CEO Fateen Anam Rafid said in his statement. "We have communicated with all active users about our security protocols and continue to strengthen our systems with new protections," as well as hiring a security firm to validate its investigation and monitor its systems in the future.

Read more of this story at Slashdot.

Google is letting outsiders experiment with DeepMind's Genie 3 "world model" via Project Genie, a tool for generating short, interactive AI worlds. The caveat: it requires a $250/month AI Ultra subscription, is U.S.-only, and has tight limits that make it more of a tech demo than a game engine. Engadget reports: At launch, Project Genie offers three different modes of interaction: World Sketching, exploration and remixing. The first sees Google's Nano Banana Pro model generating the source image Genie 3 will use to create the world you will later explore. At this stage, you can describe your character, define the camera perspective -- be it first-person, third-person or isometric -- and how you want to explore the world Genie 3 is about to generate. Before you can jump into the model's creation, Nano Banana Pro will "sketch" what you're about to see so you can make tweaks. It's also possible to write your own prompts for worlds others have used Genie to generate. One thing to keep in mind is that Genie 3 is not a game engine. While its outputs can look game-like, and it can simulate physical interactions, there aren't traditional game mechanics here. Generations are also limited to 60 seconds, as is the presentation, which is capped at 24 frames per second and 720p.

Read more of this story at Slashdot.

NVIDIA has officially launched a native GeForce NOW client for Linux as a Flatpak, giving Linux gamers access to cloud-rendered RTX gaming. Phoronix reports: While confined to a Flatpak, for now NVIDIA is just "officially" supporting it on Ubuntu 24.04 LTS and later. Granted, thanks to Flatpak it should run on other non-Ubuntu distributions too but in terms of the official support and where they are qualifying their builds they are limiting it just to Ubuntu 24.04 LTS and later. [...] At launch the Flatpak build is also just for x86_64 Linux with no AArch64 Linux builds or similar at this time. Running GeForce NOW on Linux while games are rendered in NVIDIA's cloud with Blackwell GPUs, you still need to be using a modern GPU with H.264 or H.265 Vulkan Video support NVIDIA isn't yet supporting Vulkan Video AV1 with GeForce NOW on Linux but just H.264/H.265. If you are using NVIDIA graphics the NVIDIA R580 series or newer is recommended while using the X.Org session. If you are using Intel or AMD Radeon graphics, Mesa 24.2+ is recommended and using the Wayland session. When you are up and running with GeForce NOW on Linux, you have access to over 4,500 games. The free tier of GeForce NOW provides standard access to the gaming servers and limited session caps for an introductory-level experience. It's with the performance tier where you can enjoy RTX ray-tracing and 1440p @ 60 FPS performance and up to six hour sessions. With GeForce NOW's Ultimate tier is where you are running on GeForce RTX 5080 GPU servers with support for up to 5K @ 120 FPS gaming or 1080p @ 360 FPS with up to eight hour gaming sessions in length.

Read more of this story at Slashdot.

An anonymous reader quotes a report from Ars Technica, written by Dan Goodin: Two security professionals who were arrested in 2019 after performing an authorized security assessment of a county courthouse in Iowa will receive $600,000 to settle a lawsuit they brought alleging wrongful arrest and defamation. The case was brought by Gary DeMercurio and Justin Wynn, two penetration testers who at the time were employed by Colorado-based security firm Coalfire Labs. The men had written authorization from the Iowa Judicial Branch to conduct "red-team" exercises, meaning attempted security breaches that mimic techniques used by criminal hackers or burglars. The objective of such exercises is to test the resilience of existing defenses using the types of real-world attacks the defenses are designed to repel. The rules of engagement for this exercise explicitly permitted "physical attacks," including "lockpicking," against judicial branch buildings so long as they didn't cause significant damage. [...] DeMercurio and Wynn's engagement at the Dallas County Courthouse on September 11, 2019, had been routine. A little after midnight, after finding a side door to the courthouse unlocked, the men closed it and let it lock. They then slipped a makeshift tool through a crack in the door and tripped the locking mechanism. After gaining entry, the pentesters tripped an alarm alerting authorities. Within minutes, deputies arrived and confronted the two intruders. DeMercurio and Wynn produced an authorization letter -- known as a "get out of jail free card" in pen-testing circles. After a deputy called one or more of the state court officials listed in the letter and got confirmation it was legit, the deputies said they were satisfied the men were authorized to be in the building. DeMercurio and Wynn spent the next 10 or 20 minutes telling what their attorney in a court document called "war stories" to deputies who had asked about the type of work they do. When Sheriff Leonard arrived, the tone suddenly changed. He said the Dallas County Courthouse was under his jurisdiction and he hadn't authorized any such intrusion. Leonard had the men arrested, and in the days and weeks to come, he made numerous remarks alleging the men violated the law. A couple months after the incident, he told me that surveillance video from that night showed "they were crouched down like turkeys peeking over the balcony" when deputies were responding. I published a much more detailed account of the event here. Eventually, all charges were dismissed.

Read more of this story at Slashdot.

The preprint repository arXiv will require all submissions to be written in English or accompanied by a full English translation starting February 11, a policy change that explicitly permits the use of AI translators even as research suggests large language models remain inconsistent at the task. Until now, authors only needed to submit an abstract in English. ArXiv hosts nearly 3 million preprints and receives more than 20,000 submissions monthly, though just 1% are in languages other than English. Ralph Wijers, chair of arXiv's editorial advisory council, advises authors to verify any AI-generated translations. "Our own experience is that AI translation is good but not good enough," he says. A 2025 study from ByteDance Seed and Peking University ranked 20 LLMs on translation quality between Chinese and English; GPT-5-high scored nearly 77, just below the human expert benchmark of 80, but most models including GPT-4o, Claude 4, and Deepseek-V3 scored under 60.

Read more of this story at Slashdot.

An anonymous reader shares a report: The US is leading a huge global surge in new gas-fired power generation that will cause a major leap in planet-heating emissions, with this record boom driven by the expansion of energy-hungry datacenters to service AI, according to a new forecast. This year is set to shatter the annual record for new gas power additions around the world, with projects in development expected to grow existing global gas capacity by nearly 50%, a report by Global Energy Monitor (GEM) found. The US is at the forefront of a global push for gas that is set to escalate over the next five years, after tripling its planned gas-fired capacity in 2025. Much of this new capacity will be devoted to the vast electricity needs of AI, with a third of the 252 gigawatts of gas power in development set to be situated on site at datacenters. All of this new gas energy is set to come at a significant cost to the climate, amid ongoing warnings from scientists that fossil fuels must be rapidly phased out to avoid disastrous global heating.

Read more of this story at Slashdot.

An anonymous reader shares a report: U.S. life expectancy rose to a record high of 79 years in 2024, an increase of six months from the previous year, reflecting a sharp decline in deaths from COVID-19 and drug overdoses, the Centers for Disease Control and Prevention said on Thursday. According to a report from the CDC's National Center for Health Statistics, life expectancy improved for both men and women across races and among Hispanics, surpassing the previous peak set in 2014.

Read more of this story at Slashdot.

Microsoft wants you to know that it knows that Windows 11, now used by a billion users, has been testing your patience and announced that its engineers are being redirected to urgently address the operating system's performance and reliability problems through an internal process the company calls "swarming." "The feedback we're receiving from our community of passionate customers and Windows Insiders has been clear. We need to improve Windows in ways that are meaningful for people," Pavan Davuluri, president of Windows and devices, told The Verge. The company plans to spend the rest of 2026 focusing on pain points including system performance, reliability, and overall user experience. January has been particularly rough for Windows 11. Microsoft issued an emergency out-of-band update to fix shutdown issues on some machines, then released a second out-of-band fix a week later to address OneDrive and Dropbox crashes. Some business PCs are also failing to boot after the January update because they were left in an "improper state" after December's monthly update failed to install. Users have also grown frustrated by aggressive Edge and Bing prompts, constant OneDrive upselling nags, and Microsoft's push to require Microsoft accounts. The core members of the company's Windows Insider team recently moved to different roles. "Trust is earned over time and we are committed to building it back with the Windows community," Davuluri said.

Read more of this story at Slashdot.

The private equity industry is experiencing a quiet reckoning as hundreds of midsize firms find themselves trapped between investors who have lost patience and portfolios of companies they cannot sell at acceptable prices. "There is existential risk for a number [of funds] because of the fundraising environment," said Sunaina Sinha Haldea, global head of private capital advisory at Raymond James. "If existing investors don't come and support them, new investors are highly unlikely to." According to data from Preqin, the average buyout fund that closed in 2025 spent 23 months fundraising, up from 16 months in 2021, and the total number of funds raised fell to 1,191 from 2,679 over the same period. New York's Vestar Capital scrapped plans for its eighth fund in late 2024 and has not invested in a new portfolio company since 2023. The firm's assets under management dropped from $7 billion fifteen years ago to $3.3 billion in 2024. Three-year annualized returns through June 2025 for the Cambridge Associates U.S. Private Equity Index stand at 7.4%, trailing the MSCI World stock index by 11 percentage points annually. The average holding period for buyout deals has stretched to 6.3 years from 5.1 years in 2020. Blue-chip megafunds continue raising capital normally, but smaller firms face existential pressure.

Read more of this story at Slashdot.

Apple has acquired Q.AI, a secretive Israeli startup whose technology can analyze facial skin micro-movements to interpret "silent speech," in a deal valued at close to $2 billion that marks the iPhone maker's second-largest acquisition ever, according to backer GV (formerly Google Ventures). The four-year-old company was founded in Tel Aviv in 2022 by Aviad Maizels, Yonatan Wexler and Avi Barliya. Patents filed by Q.AI show its technology being deployed in headphones or smart glasses to enable non-verbal communication with an AI assistant. The acquisition comes as Meta's Ray-Ban smart glasses already let wearers talk to its AI, and Google and Snap are preparing to launch competing devices later this year.

Read more of this story at Slashdot.

An anonymous reader shares a report: Chat & Ask AI, one of the most popular AI apps on the Google Play and Apple App stores that claims more than 50 million users, left hundreds of millions of those users' private messages with the app's chatbot exposed, according to an independent security researcher and emails viewed by 404 Media. The exposed chats showed users asked the app "How do I painlessly kill myself," to write suicide notes, "how to make meth," and how to hack various apps. The exposed data was discovered by an independent security researcher who goes by Harry. The issue is a misconfiguration in the app's usage of the mobile app development platform Google Firebase, which by default makes it easy for anyone to make themselves an "authenticated" user who can access the app's backend storage where in many instances user data is stored. Harry said that he had access to 300 million messages from more than 25 million users in the exposed database, and that he extracted and analyzed a sample of 60,000 users and a million messages. The database contained user files with a complete history of their chats with the AI, timestamps of those chats, the name they gave the app's chatbot, how they configured the model, and which specific model they used. Chat & Ask AI is a "wrapper" that plugs into various large language models from bigger companies users can choose from, Including OpenAI's ChatGPT, Anthropic's Claude, and Google's Gemini.

Read more of this story at Slashdot.

Microsoft's Xbox hardware revenue fell 32% in the final quarter of 2025 and overall gaming revenue declined 9% year-over-year, according to the company's latest quarterly earnings, released as part of results showing Microsoft's total revenue exceeded $80 billion. Xbox content and services revenue, which includes Game Pass, dropped 5%.

Read more of this story at Slashdot.