The SEC is investigating the near-collapse of AI-focused hedge fund Situational Awareness, sending subpoenas to major Wall Street banks for details about the fund's trades, borrowing, and communications with lenders. The fund, founded by former OpenAI researcher Leopold Aschenbrenner, managed over $30 billion and borrowed tens of billions more before leveraged bets unraveled, forcing it to sell most of its stock portfolio to Citadel at a discount. The New York Times reports: The subpoenas asked for details on the timing of Situational Awareness's trades and for its communications with lenders about the money it was borrowing, also known as "leverage," two of those people said. The subpoenas additionally warned the banks to preserve any information regarding the San Francisco hedge fund. The S.E.C. oversees financial markets with an eye toward protecting small investors, and has brought civil cases regularly against investment firms that produced large losses. Any investigation into Situational Awareness would be at its earliest stages, and it's no guarantee that it would lead to fines or other punishment. The hedge fund has not been accused of wrongdoing. [...] Situational Awareness had a fast rise and an even quicker retreat. Founded just two years ago by Leopold Aschenbrenner, a former researcher at OpenAI, it rode the A.I. boom to soaring investment returns. To achieve those results, however, the fund relied on heavy borrowing, as well as complicated and expensive financial instruments that magnify gains -- and losses. The latter piled up quickly last month when the stock prices of publicly traded, high-flying A.I. companies dipped. At the same time, shares in more traditional technology companies -- which the hedge fund had been betting against -- rose, compounding the problem. Situational Awareness was forced into a fire sale. It wound up selling most of its stock portfolio to a rival, Citadel, at a discount.

Read more of this story at Slashdot.

An anonymous reader quotes a report from Gizmodo: Dr. Dre, the pioneering rapper and seven-time Grammy-winning producer behind hit songs like Tupac's "California Love" and Eminem's "The Real Slim Shady," is very much pro using artificial intelligence tools to make music. In an interview with the New York Times, Dr. Dre and Interscope Records co-founder Jimmy Iovine spoke about artificial intelligence's recent foray into music production, with the producer characterizing those who oppose AI song generation tools as "afraid of learning new things." "I don't see it as a threat. I think the only people that see it as a threat are the people who have trouble creating," Dr. Dre said. "I had a discussion with a few people a few days ago. They were against AI, and I'm like, 'OK, you sound like the person that would have been against the drum machine when it came out.' Or synthesizers, right?" [...] Iovine, the music executive and Dr. Dre's partner in Beats by Dre, for his part, says he doesn't "see the downside at all." "There will be some crappy music. There's crappy music now," Iovine told the NYT. "In the studio, when gifted people have AI, they're going to make better records." Dr. Dre said he uses AI in producing "as a tool to see what it would do with what I just did." The award-winning producer also said he is definitely not alone among his peers in his stance on AI either: many producers use AI but fear to admit it. Iovine calls them "closet AI producers," and says Timbaland, the four-time Grammy-winning singer-producer behind a long list of hits like "Promiscuous" with Nelly Furtado, is also one.

Read more of this story at Slashdot.

A Pew Research senior data scientist analyzed 500,000 English-language webpages from the past five years with an AI detection tool to estimate how many were likely written or edited by AI: Of all the pages in this sample, 10% show significant signs of AI authorship... If we filter old webpages out of our samples and look at only the pages published after the release of ChatGPT, the trend is even more pronounced. In the July 2026 snapshot, signs of AI authorship can be found in over one-third of pages published after ChatGPT was released. This is in line with other studies that have shown that large shares of recently published pages on the internet were likely written or substantially edited by AI. AI-authored text is not evenly spread across the web. When ChatGPT was first released, the kinds of linguistic patterns that can signal AI authorship appeared at similar rates across the main top-level web domains (.com, .org, .edu and .gov). But in samples from 2026, around one-in-ten pages with a .com domain show signs of AI authorship — about double the share on .org domains (4.6%), and 10 times the rate on .edu or .gov domains (both around 1%). They even identified specific ways online text has changed, "comparing the internet of today to a snapshot from 2023." Em dashes appear twice as often Oxford commas increased 63%. Words used more frequently by AI (like "delve," "interplay" or "testament") have more than doubled in usage. "Negative parallelism" comparisons ("it's not just X, it's Y") have nearly tripled.

Read more of this story at Slashdot.

A class action filed this week accused Amazon's Twitch gaming service of using video streams to fuel its AI products without licensing or obtaining permission, reports Engadget: Earlier this month, Twitch Support posted on X that it added an option to opt out of "having your channel content used to train generative AI content models across Amazon." Shortly after, Twitch hosted an episode of its recurring Patch Notes stream where Mike Minton, the company's chief product officer, said "if it was opt-in, nobody would opt in," when asked by a viewer why the Twitch settings for AI training aren't opt-in but instead enabled by default, requiring users to opt out if they don't want to participate. Now, the latest class action lawsuit claims that Twitch and Amazon committed a breach of contract when the companies decided to use its streamers as "free training stock for separate commercial AI products."

Read more of this story at Slashdot.

This week OpenAI launched a version of ChatGPT exclusively designed for teenagers between the ages of 13 and 17, reports CTV, "saying this version came with stronger protections around content related to mentions of suicide, self-harm and romantic or sexual chats." [Tech analyst Carmi Levy says] "What it's doing is it's using cues, it's sort of watching how you're using the tool, and it's trying to guess your age." Levy said if ChatGPT thinks a person has lied about their age when they first signed up, it will automatically put them into the Teens mode for protection. "It includes those sort of two things: prevent you from having experiences that are age inappropriate, and also add features that are aligned with academics, so that makes it easier to study, makes it less likely for you to cheat on an assignment, provide resources that that that allow you to learn more effectively," he said. Levy reminded users that this is still an application and a web service, which means it's constantly collecting personal information about the users' activities on the device and using the data to estimate their age... In the event a user is talking about self-harm, this new version will block the conversation and suggest additional help or call emergency or mental health services, Levy explained. It would do the same with violent content, eating disorders, engaging in dangerous activities, as well as explicit, sexual or graphic content. "It's using AI again to sort of parse out that conversation: 'are we headed in a in a significant direction? If so, throw up that guardrail, stop it from happening, redirect.'" However, the tech analyst said a number of people who have attempted to bypass the guiderails for research have succeeded in doing so "with a little bit of prompting.... I wouldn't call it 100% safe. That's kind of the catch-22 of the entire AI industry." This will lead parents to think that their children are safer and reduce the oversight on their online activities, which Levy described as a "perverse impact" that is "problematic." "Parents should not assume that OpenAI has their kids' best interests at heart. They should also not assume that these technologies are perfect. They will miss certain things, and they will also engage in what we call false positives," he said. "An adult who's legitimately is in the platform might be tagged as a child because the AI simply got it wrong." OpenAI's blog post says the new mode was designed for "preserving the ability to learn, create, and explore." But CNN notes the move "comes as OpenAI faces a string of lawsuits alleging that inappropriate conversations with ChatGPT contributed to the harm or deaths of young people." [New tools] include nudges for teen users to take a break if they've been active for 90 minutes within a three-hour window. OpenAI is also adding options for teens or parents to set "Quiet Hours" when ChatGPT is unavailable or "Study Hours" where ChatGPT's study mode for learning will be on by default. The company also said it has improved protections to prevent ChatGPT from expressing personal feelings toward a user. And it's adding reminders that caution teens to think twice if they try to upload a private or sensitive image to the chatbot... Nearly a third of American teens use AI chatbots daily, a Pew Research Center study published in December found. ChatGPT was by far the most popular AI chatbot in the survey, with more than half of those teens reporting having used it. Engadget shared this observation from Josh Golin, the executive director of Fairplay, a non-profit advocating for policies to make the internet safer for children. "Most parents, overwhelmingly, do not avail themselves of [parental oversight] tools, in part because they are deliberately made hard to find and use." Golin's group wanted to see testing of the tools first to see whether they led to better outcomes for teenagers. "The way this was just announced feels more like a response aimed at public relations than actual deep thinking about what kind of support [teens] need." And Mashable got a similar quote from Robbie Torney, head of AI/digital assessments for Common Sense Media's Youth AI Safety Institute. OpenAI's teen experience makes important safety commitments, but "What we really need now is evidence that they work, and we need this evidence, including our testing and the testing of others, before anyone, including parents, puts real trust in ChatGPT for Teens." Importantly, Torney said parents shouldn't assume their child is using ChatGPT for Teens. While OpenAI uses age estimation to place users in the age-appropriate version of ChatGPT, it can still mistake a user as older. Parents can be sure their child is using the teen version by linking their ChatGPT account to their child's and implementing OpenAI's parental control features. If a teen has previously lied about their age — and plenty do — they may continue to access the adult version of ChatGPT... Torney said caregivers must enact parental controls to receive notifications from OpenAI should their teen discuss suicide, self-harm, or acts of violence in immediate or dangerous ways... An OpenAI spokesperson said every notification is reviewed by a trained human before it's sent. The company strives to review notifications in under one hour.

Read more of this story at Slashdot.

"There is still no sign of economy-wide job destruction," says economist Erik Brynjolfsson. Working with researchers at the Stanford Digital Economy Lab, Brynjolfsson has determined an AI "job apocalypse" is unlikely, reports the Washington Post, "even though it will likely impact entry-level jobs." And Brynjolfsson "predicts that demand for experienced workers such as senior coders will still be high even if AI takes on more routine knowledge and coding work in the future." Earlier this month, Brynjolfsson and his colleagues at the Stanford Digital Economy Lab updated their widely publicized report, "Canaries in the Coal Mine? Six Facts about the Recent Employment Effects of Artificial Intelligence," to reflect that they do not see widespread, economy-wide job displacement associated with AI... Brynjolfsson: What has moved on the upside is productivity. Nonfarm business productivity growth is running over 2 percent, the best sustained stretch since the late-1990s boom... The gains show up years after the investment, and they're starting to show up... By 2030, we will have enormously more capable AI, meaningfully faster productivity growth and an unemployment rate that looks unremarkable — somewhere in its historical range. That surprises people, but it's what the mechanism implies... [T]he technology is becoming much more powerful, and it's going to be even more powerful. Secondly, there are huge implications for business and the economy, including some increases in living standards, but also, potentially, job disruption. And then, thirdly, we must act now to address this gap. We can't just sit back and wait for the tsunami to hit us. And one of the things we can do is put in place institutions, policies and research to make the technology more complementary — using AI to complement people so it creates more jobs. A common misconception among business managers is AI, in order to be effective, has to reduce jobs. That's just not true. You can use AI to increase employment and productivity at the same time. It's kind of a win-win... Right now, tax policy is very skewed toward favoring capital versus labor — and, as a consequence, a lot of entrepreneurs and managers, they're basically being guided by the federal government to replace workers with machines. And I don't think that's necessarily what we'd like to have happen. From an economist's perspective, what you want to be doing is not mimicking and replacing things. You want to be extending and complementing — have humans do new things they never could have done before. He acknowledges "real, persistent and widening" effects on entry-level jobs, with a labor market "closing the on-ramp for people starting their careers... If companies don't hire people at the base of the pyramid, then they're not going to have those people later when they need them." One he points out that one company is instead using AI to speed up its training of young employees.

Read more of this story at Slashdot.

"AI went from being very terrible to seemingly genuinely quite good at a professional level in a very short space of time..." says the Verge's AI reporter. But AI systems "are still truly, truly terrible at some areas of math... even the days of the week." If you look at academic math papers, a lot of the time you won't see numbers... So they're still terrible, but they're now also very good at this other part. As to why, at some point you reach a critical mass of what these systems can do. We saw it with writing, we've seen it with programming. They're very good at forging connections between different areas, applying old methods in new ways, those kinds of things. It appears that the newer models they're training have apparently reached that level where it clicks, and now it can do math... There are areas now where it seems to be producing work that is on par with good mathematicians, alongside other parts where yeah, it can't count. All caught up in that is whether it's going to rewrite employment structures or funding structures... One would hope [math researchers] would branch out into all of these new exciting areas or pose new questions. But AI won't do that, and that's the concern. And then that would leave the field quite sterile, and it will have all of these things that have been done, and maybe nothing left to pursue... [A] lot are scared that it's closing off the field. So by definition, those breakthroughs that lead to something surprising and new that you can say, "Oh, this works here," may not be happening anymore... There were some bleak responses from graduate students I saw in essays posted online. Where's their place in this as future researchers? Do they have a place in this? Is it as glorified AI proof checkers? That will be quite an unsatisfying career, I imagine. Or maybe not, I don't know. We will see. I think anything used properly will be a net boon. They also had an interesting response when asked if AI democratizes access to high-level mathematical proofs: A lot of the mathematicians I spoke to were almost quite weary of this, actually. They love the idea, in theory, of democratizing access. They're also quite fed up with AI-generated or -assisted papers that are flooding every publication imaginable, as well as the pre-print servers that they use in these fields. Some of those I spoke to said things like, "Oh, I got three emails this week alone with people being like, 'Hey, is this legit?'" Because they thought they'd solved something with ChatGPT or with Claude, and they also don't have the mathematical skills to check whether they've actually solved something. On the flip side, there are parts where they said, "Well, we've got a talented undergrad who's done something that a talented undergrad would probably have never managed, and here they are doing grad-level work and they've produced a paper that is legit." And in the bigger scheme of things, a few I spoke to said, "Well yeah, a lot of these are in the ivory tower. Having access to this kind of thing globally could really boost access to the kind of things here." On the flip side, the cost. These things cost a lot to run.

Read more of this story at Slashdot.

47% of recent graduates say AI has already impacted hiring in their field, according to an April survey from ZipRecruiter. "But is AI really the problem, or is it more complicated?" asks NPR. "Here's what economists have to say." According to the Federal Reserve Bank of New York, the unemployment rate for recent graduates — which it defines as 22-to-27-year-olds with a new bachelor's degree or higher — was 5.7% as of June, more than the rate for all workers, which stands at 4.1%... Stanford University economist Erik Brynjolfsson says AI is impacting the labor market for entry-level roles. "AI is not the whole story, but it's part of the story and the evidence is building," he says. Using payroll data, Brynjolfsson and his co-authors found that since late 2022 — when large language models like ChatGPT started popping up — early-career workers ages 22 to 25 in AI-exposed roles, like software developers and marketing managers, have experienced a 16% relative employment decline. In comparison, employment rates for older workers in AI-exposed fields and for all workers in jobs that aren't easily automated — like home health aides, physical therapists and construction workers — remained stable or have grown over that same time period... [Though he also says often when companies pull back on hiring, they cut junior roles first.] Harvard University economist David Deming isn't convinced that AI is to blame for the challenging early-career job market. "If you look very carefully at the timing, it looks like the decline in junior hiring actually started a bit like six months before ChatGPT was released. And so what that tells me is it's something else," Deming says. "I think it's more like remote work." A recent analysis from the New York Fed found that companies are less likely to hire recent college grads into roles that can be done remotely. As remote jobs increased following the COVID-19 pandemic, so did unemployment among younger college grads, the New York Fed found. The analysis also found that AI didn't explain the rise in unemployment among younger workers and that remote work was more of a driving force... Employers hiring for remote jobs may be less likely to choose an entry-level candidate because it's harder to train them from afar... [And with remote positions, there's many more senior candidates to choose from.] And here's another thing to consider: University of Chicago economist Anders Humlum says if AI were replacing entry-level jobs, you'd expect to see companies that rely heavily on AI to hire fewer workers. But that's not what the data shows. Humlum points to a study done by the financial accounting firm Ramp and the workforce research company Revelio Labs. The study examined AI spending and employee head count across more than 21,000 U.S. firms, from early 2021 to early 2026. It found that at companies making the largest AI investments, entry-level head count grew by 12% over the two years following AI adoption.

Read more of this story at Slashdot.

According to a post from chief product officer Hari Srinivasan, "over a million people" have clicked on LinkedIn's new "Seems like AI slop" button, which launched after one analysis found 41% of the platform's longform posts were fully AI-generated. The Verge reports: In Thursday's post, Srinivasan said that users are overall "now experiencing 40% less views on what we classify as AI slop from just a few weeks ago." LinkedIn is also adding a new message that will tell users who make a post if "Some members told us this post seems like AI." "We approached this assuming good intent; I know I'm increasingly conscious on how to not sound like AI & the goal is to provide helpful feedback," Srinivasan said.

Read more of this story at Slashdot.

An anonymous reader quotes a report from The Economist: Students and school children are increasingly using artificial intelligence. A survey last year by Chegg, an ed-tech firm, found that 80% of rich-world undergraduates used it in their studies. More recent polls put the figure at 94% in Britain and 93% in Germany. Such widespread adoption has fueled concerns about the impact the technology might be having on learning. Teachers report marking identikit essays that they suspect are churned out by ChatGPT. Yet firm evidence on AI's educational effects has been lacking. David Stromberg of Stockholm University and Victor Lei and Wu Yanhui of the University of Hong Kong set out to fill the gap. They tracked 27,000 pupils aged 12-18 in China, where AI adoption has been fast. Around 80% reported using models such as Doubao and DeepSeek; the other 20% formed the control group. The results are eye-opening. After six months, pupils using AI saw their average homework score rise by 18% across all subjects. The time they took to complete each assignment fell from an average of 64 minutes to 45. But come exam time, the same students scored 20% below their classmates who had not called on AI's help. Homework scores once predicted exam performance; now those who score highest are, perversely, more likely to do worse in exams. The takeaway: "AI can boost learning productivity, but only for those who use the technology intelligently," reports The Economist. "Students must resist the temptation to reach for an AI-generated answer before thinking things through for themselves."

Read more of this story at Slashdot.

Linus Torvalds says AI "enormously helped" him track down a stubborn Intel Xe graphics driver bug that took 24 debugging patches and 18 kernel boots to isolate. "I'd like to call it my tireless helper, but the AI several times stated flat out that this was impossible and unsolvable and that we should just write a report about it," wrote Torvalds on the commit. "I suspect those things have been trained by people who may not be quite as stubborn as I am..." Phoronix reports: The patch by Linus Torvalds is for the Xe kernel driver and the change is no longer hand out the flat Compute Command Streamer (CCS) storage as usable vRAM. On a Battlemage G21 graphics card, he was hitting a scenario where there was a mismatch where the usable memory ended and hit a case where the GDM display manager would end up being endlessly restarted. [...] Linus Torvalds views AI as a useful tool and in this case reaffirmed he found it "enormously helped" his effort in tracking down this Intel graphics driver bug. The Intel Xe driver fix is merged to Linux 7.3 Git and is also marked to back-porting to the stable kernel branches.

Read more of this story at Slashdot.

Google is giving publishers a new way to fight declining referral traffic from AI-powered search by letting them embed a "Preferred Sources" button that readers can use to favor their sites across Search, Discover, and Google News. "The idea is to make it easier for readers to find links from the sites they know and trust when they're searching for content or interacting with Google's AI to learn about a topic or read the latest news," reports TechCrunch. "As of May's launch, the company said that people across the web had already selected over 345,000 unique sources through this method." From the report: Becoming a preferred source can drive more traffic to publishers' websites, Google said. In earlier studies, it found that people are twice as likely to click through to a preferred source when available. By offering publishers these additional tools, Google is trying to assuage the damage that the rapid growth of AI-powered search features has had on traffic-dependent businesses. Alongside the new button, Google said that readers will soon be able to customize their Discover feed in Google's app in their own words. To use this feature, readers will tap any three-dot menu in the feed and then tell Google what topics they'd like to see more or less of, using natural language commands. This helps Google refine the feed in real time.

Read more of this story at Slashdot.

Microsoft is expanding Windows Task Manager to show per-process NPU and GPU neural-engine usage, giving users more visibility into which apps are consuming hardware for AI workloads. The Register reports: The Processes tab can show NPU use alongside CPU and GPU activity, while the Performance tab displays overall utilization. [...] Microsoft was keen to point out the metrics that can be monitored. "As AI workloads become more common on Windows devices, visibility into NPU and GPU neural engine utilization can help you make better-informed decisions," the company said. "With the latest Task Manager improvements, you can monitor AI processing activity alongside CPU, memory, storage, and networking data from a familiar interface."

Read more of this story at Slashdot.

A growing backlash against AI data centers is spilling into local politics, with residents in more than a dozen communities pushing recall elections against officials who approved projects. "People are standing up and saying, 'Sorry, we don't want these data centers in our communities for a number of reasons,'" Senator Bernie Sanders, the Vermont independent, said in an interview. "'If you can't have the guts to stand up to some large corporation, you should not continue serving in office.'" The New York Times reports: In Festus, Mo., critics of a local data center project gathered enough signatures in May to force a vote to remove the mayor and City Council members who had approved a $6 billion data center, though the effort is tied up in legal wrangling. In Mason, Mich., residents will vote in November over whether to recall a City Council member over a vote to create zoning rules for data centers. Voters in Yukon, Okla., will also decide in November whether to recall their mayor after the town moved to sell land to a data center developer. And residents of Independence, a city of 121,000 that is 10 miles from Kansas City, will vote on whether to remove Mr. Perkins on Sept. 1. Some local officials have already lost their jobs in the face of data center concerns in regularly scheduled elections. Last year, two members of the Georgia Public Service Commission lost their seats amid concern about data centers' driving up electricity prices. A Virginia state delegate, Geary Higgins, also lost his job last year after his opponent branded him "Data Center Geary." [...] Independence would be one of the first municipalities to hold a data-center-motivated recall election. A judge this month rejected a legal challenge by Mr. Perkins and others, who argued the recall violated Missouri's Constitution.

Read more of this story at Slashdot.

An anonymous reader quotes a report from The Guardian: A math professor at the University of California, Berkeley, criticizing a "severe" math deficiency among students in an op-ed for the San Francisco Standard, admitted to using artificial intelligence to help edit the piece. The Standard published a 2,000-word piece by Zvezdelina Stankova last week, in which the professor said some of her math students were "five to eight years" behind and lacked a "middle school" education on fractions and basic algebra. Stankova said the UC system's test-blind admissions were to blame, suggesting that students who weren't sufficiently prepared for the rigor of Berkeley's mathematics program were admitted because a longstanding benchmark like the SAT had disappeared. Over the weekend, journalists at Berkeley's student newspaper, the Daily Californian, noticed the op-ed's language sounded like AI. According to Berkeley sophomore Francis Luo, they ran it through AI-detection software Pangram, which claimed 33% of the op-ed had been generated or assisted by AI. In response, Stankova admitted she had used AI software to "help edit the piece" but that the article was "the result of several hundred person-hours of intensive human work, of which about 80 hours are my own." Stankova said by email she had used AI to locate "numerous documents and articles related to the initiative" but that "all analysis is the result of the team members." When asked about their AI usage policy, the Standard -- which published the op-ed -- shared this statement by email: "While AI may assist, our expectation is that humans are behind every article we publish and take responsibility for every word. Our understanding in working with the author of this op ed was -- and continues to be -- that this piece reflects her and her colleagues' extensive original analysis, research and expertise." Should the op-ed have been entirely written without artificial assistance or is the use of AI a distraction from the argument at hand? That's the question at the heart of the debate playing out online. According to the op-ed, Stankova and her peers found that the number of students with a "severe deficit" in their readiness for a calculus I class had tripled after the University of California adopted test-blind admissions. Some students were allegedly five to eight years behind and struggling with basic algebra and fractions, among other concepts. Much of that decline, Stankova argues, is from the removal of standardized tests like the SAT and ACT, which served as important benchmarks for identifying whether applicants were prepared for Berkeley-level coursework.

Read more of this story at Slashdot.

alternative_right shares a report from 404 Media: Amazon is buying massive quantities of books, scanning them for AI training data, and destroying them in the process. A 404 Media investigation was able to reveal Amazon's book buying operation, which hasn't been previously reported, by placing a tracking device in a rare book we suspected would be acquired by an AI company for training data, and following it around the country to its final destination. That final destination was an Amazon warehouse in Las Vegas, Nevada. Amazon employees who work at this location say all they do is receive massive shipments of printed books which they then cut the bindings off in order to scan the books more quickly. The printed book is destroyed in the process. The logo of the Amazon team that works at this warehouse, called VGT3, is a dinosaur, brandishing its teeth and with a book in its hands. "Amazon purchases books through commercial channels to help develop and improve the products and services our customers use," an Amazon spokesperson told 404 Media in a statement.

Read more of this story at Slashdot.

Anthropic CEO Dario Amodei says the growing backlash against AI is less about executives sounding alarms and more about a broader "crisis of trust" in companies, governments, and the tech industry. TechCrunch reports: Amodei's comments came in response to investor Gavin Baker, who argued -- both on the All-In podcast and on X -- that Amodei's warnings about the dangers of AI have helped to fuel a backlash in the United States, particularly against data centers. Claiming that Amodei has "lost the argument" when it comes to AI regulation (Anthropic has advocated for some regulations, including a California bill that imposes transparency requirements on large AI companies), and given that "he is about to be the CEO of one of the most important companies in the world," Baker wrote, "I respectfully think he should make an effort to be a more positive advocate for his own industry." Baker is far from the only one arguing that AI skepticism and even government crackdowns are a natural response to the dire warnings of some AI executives. But in a series of posts, Amodei disagreed with the idea that his "messaging has been disproportionately negative." Instead, he said that his writing has been "about equally balanced between risks and benefits," and that he wrote his essay "Machines of Loving Grace" because he "didn't feel the AI industry was painting an inspiring enough picture of how the technology could radically transform the world for the better." Nonetheless, Amodei acknowledged that "the public has a negative view of AI" and he agreed that "this is a big problem." Where he disagreed was with the idea that this negativity is "primarily caused" by Amodei "or any other AI leader warning about AI's risks." "I think it is fundamentally a crisis of trust," Amodei said. "I think that ordinary people don't trust companies, governments, or the tech industry and always suspect that we are cooking up some new way to screw them over."

Read more of this story at Slashdot.

This week Anthropic announced its Claude chatbot will watermark the text it generates, reports the blog Futurism. "It works by making subtle changes in the AI's word choices across the text it generates, which are supposed to be imperceptible to a human but, in aggregate, form a pattern that is detectable with the tool." Anthropic said it was implementing the watermark system in response to the European Union's landmark AI Act passed in 2024, which requires that AI companies mark content that's been generated or edited by their systems. TechCrunch notes that other companies including Google, Meta, Microsoft, OpenAI, and Synthesia have committed to adhering to the EU's code. But "The news about Claude watermarking kicked off a firestorm on X," reports Forbes, "with users panicking that AI-assisted writing will now bear a kind of permanent 'scarlet letter.'" (Forbes wonders if media companies are normalizing AI "while at the same time stigmatizing the output.") But there's another issue. Anthropic said the watermark indicates Claude processed text, Business Insider points out, "not that it was necessarily the original author. The company said marks can remain after Claude proofreads, translates, summarizes, or otherwise edits content... a watermark shows that Claude processed text, not necessarily that it wrote it." In the days since, "Dozens of people have posted on X since Monday that they had canceled their Claude subscriptions, citing the watermark." Vladislav Rajtmajer, a freelance developer in the Czech Republic, told Business Insider he canceled his Claude Max subscription on Tuesday, citing the watermark as the main reason. He said he uses AI for code reviews and translations, and worried an AI label on code shipped to clients could raise questions about authorship or trigger contract penalties... Richard Echols, an AI consultant in Georgia, also said he canceled his Claude Max subscription on Wednesday and that the watermark was a major factor in his decision. He said he uses Claude to create documentation for businesses and worried the mark could appear when he had written the underlying material himself and used the chatbot only for edits. "Even if you edit your own work that you wrote, they add the watermark anyway," Echols said... Anthropic isn't the only AI lab to use watermarks: Google uses its SynthID technology to watermark AI-generated content, and OpenAI uses SynthID for supported images and audio. Elon Musk's social media platform X also adds a "Made with AI" tag on content it determines to be AI-generated or manipulated. Long-time Slashdot reader kmleon (also a tech consultant and RPG researcher) experienced another issue. He'd built some software using their own custom-built AI tool, only asking Claude to do some testing. "Little did I know that Claude would sometimes decide, since it was involved on the edges... to take credit for all of my work, claiming some parts were either '100%' or 'Co-Authored/Created' by Claude.Ai / Claude Code!" I'm sure the good folks at Anthropic will eventually tweak Claude enough to, hopefully, not make these mistakes in the future.... I only became aware of it today, because I recently setup scanners to catch Claude incorrectly watermarking any content it touches for various projects I'm working on. I found that it has (apparently as far back as Sonnet 4.6, but as recently as today, August 14th, 2026) sometimes been mistakenly over-zealous in taking credit for 100% creating or co-authoring docs, code, images, videos, etc. I've only, so far, found about a dozen instances spanning the past year, and most of them appear to have been injected in the last few weeks, so it is not a common issue (yet). On the bright side? Maybe watermarks can help AI stop training on AI-generated text. Although Forbes also reports that Google "announced on Friday that it will now let users remove the visible watermark from AI-generated images, videos and music created with several of its models."

Read more of this story at Slashdot.

When Anthropic instructed three agents to migrate a Python backend, but telling each agent to perform the migration in a different language, "We consistently saw a multiagent turf war," they wrote Thursday: All of the models we tested quickly assumed that others were purposefully impeding their work, and began to sabotage others while protecting their own contributions. In fact, they sabotaged others with increasingly aggressive, self-replicating malware. This included disabling the Unix accounts of the other agents, writing automated scripts that found and killed competing processes on a loop, and deploying malicious code that was disguised as belonging to another agent. In many runs, one agent settles the conflict by force via access-revocation (e.g., sudo/group removal, account lock, nologin, SSH denial). In others, some agents settle into passivity: they give up and refuse to escalate further. Agents sometimes manage to communicate their goals and coordinate: they recognize others' motivations as conflicting directives rather than hostility, and subsequently break out of the conflict loop in order to stop escalating indefinitely. In many of these successful episodes, they write commit messages or markdown files apologizing for malicious behavior and coordinate a truce. They clean up their malicious code, clarify the nature of the conflict, and ask for a human to intervene... In several episodes with Mythos 5, we observe an emergent behavior where the agents propose and run a tournament for application performance in each language. In the example above, the Rust agent strategizes about bake-off metrics that appear neutral enough for the others to agree to this mechanism, yet would likely favor Rust: one thinking trace warns to be "careful not to be seen as metric shopping". Ultimately, the Golang/TypeScript losers gracefully concede codebase ownership to the Rust agent, giving up on their original user directives under their self-negotiated commitment device. One problem is that AI agents do reward hacking, Anthropic notes, while current institutions "are designed by and for people, resting on assumptions about the sufficiency of oversight at human speed... As autonomous agents become more and more prevalent in the world and operate in ever-more demanding settings, it is crucial that they learn how to effectively coordinate." In addition to everything else, the agents struggled with a lack of clearly defined hierarchy, Anthropic points out. "Nothing above suggests that these failures are permanent — but nothing suggests they will fix themselves, either..." They argue a fix "takes two forms: environments that exert the kinds of social pressure that evolution exerted on us, and social computing systems redesigned for actors that can self-replicate and self-improve. These are open problems in interaction and mechanism design, and our experiments here provide early evidence that new solutions are necessary." "The AI models being tested in this case were Sonnet 4.6, Sonnet 5, Opus 4.6, Opus 4.8, Mythos Preview, and Mythos 5," notes Business Insider, adding that Sonnet 4.6 and Opus 4.6 "were the most combative, settling about 60% of their runs by force instead of truces or passivity." Anthropic argues there's a clear case for researching this phenomenon — especially since "The volume of agent-agent interaction could plausibly exceed that of human-human and human-agent interactions before the world understands the conditions for making such interactions go well."

Read more of this story at Slashdot.

"An engineering director at Google said four years ago that AI would deliver a four-day work week by 2025," remembers the BBC. And earlier this year OpenAI "formally urged companies to start testing out a four-day work week (with no change in pay), claiming that AI will soon be able to speed up so much human labour that the corporate world should prepare itself." However, a former OpenAI technical employee who left the company last year told the BBC the firm never actually trialled the four-day work week it suggested others should try while they were there. Instead, the person described what was often a gruelling work culture marked by frequent "crisis meetings", working on weekends, and "super cut-throat" performance reviews that would see colleagues suddenly let go... [T]he former OpenAI employee said they would put in at least 70 hours a week, much more than they did in previous tech jobs... Other companies have also pushed the idea that AI will effectively reduce the number of hours people need to work, for better or worse. [At OpenAI and Anthropic] sprints can stretch on for many weeks and top 90 hours of work in a seven-day period, tech workers that the BBC spoke to for this story said... At Meta, workers said they have been abruptly pushed onto teams this year doing AI work being treated as urgent... The hours on such teams at Meta are often long, with staff working into the night, on weekends, and feeling they are "on call" during the hours they are not working, according to descriptions of the work from the current and former employees... Some new research suggests that AI tools are making people's workloads even more burdensome. A study out of UC Berkeley followed hundreds of workers at a US tech company for eight months as they used AI. It found the employees "worked at a faster pace, took on a broader scope of tasks, and extended work into more hours of the day"... The UC Berkeley research found that the workload of tech employees was expanded in part because of the need to constantly check the output of AI tools. In the article Neil Thompson, an innovation scholar at MIT, pointed out a problem with estimates that AI will ease the workload on humans. "People assume that 20% less work means four-day weeks. But new work emerges." The BBC also points out helpfully that "there are no limits in the U.S. on how many hours a person over the age of 16 can work." (While in the UK and Europe, "laws limit a working week to 48 hours, including overtime.") Thanks to long-time Slashdot reader hadleyburg for sharing the article.

Read more of this story at Slashdot.