chatainews

"The performance of large language models (LLMs) has recently improved to the point where models can perform well on many language tasks.   "We show here that —for the first time —the models can also generate valid metalinguistic analyses of language data.  "We outline a research program where the behavioral interpretability of LLMs on these tasks is tested via prompting. LLMs are trained primarily on text —as such, evaluating their metalinguistic abilities improves our understanding of their general capabilities and sheds new light on theoretical models in linguistics.  "We show that OpenAI’s (2024) o1 vastly outperforms other models on tasks involving drawing syntactic trees and phonological generalization.  "We speculate that OpenAI o1’s unique advantage over other models may result from the model’s chain-of-thought mechanism, which mimics the structure of human reasoning used in complex cognitive tasks, such as linguistic analysis. "

"Recent results show that these models can, in principle, do sophisticated linguistic analysis. But no model has yet come up with anything original, nor has it taught us something about language we didn’t know before. "If improvement is just a matter of increasing both computational power and the training data, then Beguš thinks that language models will eventually surpass us in language skills.  "Mortensen said that current models are somewhat limited. 'They’re trained to do something very specific: given a history of tokens [or words], to predict the next token,' he said. 'They have some trouble generalizing by virtue of the way they’re trained.' "But in view of recent progress, Mortensen said he doesn’t see why language models won’t eventually demonstrate an understanding of our language that’s better than our own. 'It’s only a matter of time before we are able to build models that generalize better from less data in a way that is more creativ...

"Mark Zuckerberg’s Meta is spending untold billions on infrastructure and top talent for its AI ambitions. "In fact, the CEO announced during the company’s earnings call on Wednesday, Meta will be spending between $70 billion and $72 billion on AI this year —up from its previous estimate of $66 billion to $72 billion, as CNBC reports. "Unsurprisingly, that cash bonfire isn’t going over well with investors.  " Meta’s shares slid by more than 11 percent on Thursday , indicating widespread skepticism about the company’s ability to stop bleeding billions of dollars as it races to keep up with the AI industry’s ever-escalating expenditure commitments."

"MLPerf benchmarks have also gotten tougher. And this increased rigor is by design —the benchmarks are trying to keep pace with the industry, says David Kanter, head of MLPerf. 'The benchmarks are meant to be representative,' he says. "Intriguingly, the data show that the large language models and their precursors have been increasing in size faster than the hardware has kept up. So each time a new benchmark is introduced, the fastest training time gets longer. "Then, hardware improvements gradually bring the execution time down, only to get thwarted again by the next benchmark. Then the cycle repeats itself."

"Job postings for entry-level and early career roles are way down year over year. The market has pulled up the ladder for people trying to get in on the lower rungs, and the prospect of climbing it is getting harrowing, too. "A recent report from the Federal Reserve Bank of Philadelphia found that AI exposure is over three times higher for occupations that require a bachelor’s degree compared with those that don’t. "The idea up in the C-suite is almost certainly that automation will be able to fill in those gaps, even though there’s little to suggest that it will actually play out that way.  "According to a study done by the Center for AI Safety, AI agents were only able to complete about 3% of the work assigned to them that humans can do reliably."

"David Chalmers —Professor of Philosophy and Neural Science at New York University —defined the distinction between real and apparent consciousness at a conference in Tucson, Arizona in 1994.  "He laid out the hard problem  of working out how and why any of the complex operations of brains give rise to conscious experience, such as our emotional response when we hear a nightingale sing. "Prof Chalmers says that he is open to the possibility of the hard problem being solved. "'The ideal outcome would be one where humanity shares in this new intelligence bonanza,' he tells the BBC. 'Maybe our brains are augmented by AI systems.' "On the sci-fi implications of that, he wryly observes: 'In my profession, there is a fine line between science fiction and philosophy'."

"I've been working with the team at The Washington Post to launch the WP Incubator a new program backing founders who are building what's next for news, storytelling, and Al-driven media. WP Incubator is another milestone in the Post 's transformation journey. "As the Entrepreneur in Residence, I'm helping design the program and support the founders we'll bring in.  "Our mission is clear: to move fast, build responsibly, and create products rooted in trust, quality, and civic impact. This is not an innovation lab and we're not looking for iterative innovation. We want to build moonshots , groundbreaking ideas that could become independent ventures. "We're now open for applications so if you're building something ambitious and disruptive at the intersection of media and Al apply here (and move fast, the window is closing soon). "You'll get access to The Post 's expertise, audience insights, and network and a chance to t...

"Chatbot site Character.ai is cutting off teenagers from having conversations with virtual characters, after facing intense criticism over the kinds of interactions young people were having with online companions. "The platform, founded in 2021, is used by millions to talk to chatbots powered by artificial intelligence (AI). "But it is facing several lawsuits in the US from parents, including one over the death of a teenager, with some branding it a clear and present danger  to young people. "Now, Character.ai says from 25 November under-18s will only be able to generate content such as videos with their characters, rather than talk to them as they can currently."

"The 10 richest people have added over half a trillion dollars to their fortunes this year, almost entirely because of AI optimism, fueled by buzzy partnerships, massive contracts, and sky-high growth forecasts. "Year to date, Oracle stock has soared nearly 70%, Alphabet, Nvidia, and Dell have surged by over 40%, and Meta and Microsoft have jumped by over 25%. "All of the top 10 wealthiest people, except Arnault, are major shareholders of the world's biggest AI companies, so those share-price gains have lifted the value of their stakes and inflated their net worths. "Market watchers are divided over whether the AI boom is a bubble about to burst or the start of a tech revolution."

"Elon Musk launched Grokipedia, a fully AI-generated encyclopedia  that serves no one and nothing other than the ego of the world’s richest man. As others have already pointed out, Grokipedia seeks to be a right wing, anti-woke Wikipedia competitor. But to even call it a Wikipedia competitor is to give the half-assed project too much credit. It is not a Wikipedia competitor  at all. It is a fully robotic, heartless regurgitation machine that cynically and indiscriminately sucks up the work of humanity to serve the interests, protect the ego, amplify the viewpoints, and further enrich the world’s wealthiest man. It is a totem of what Wikipedia could and would become if you were to strip all the humans out and hand it over to a robot; in that sense, Grokipedia is a useful warning because of the constant pressure and attacks by AI slop purveyors to push AI-generated content into Wikipedia. And it is only getting attention, of course, because Elon Musk does represent an actual thr...

"Albania’s announcement that its AI minister Diella is pregnant with 83 children  is at once eccentric and ambitious. "If this experiment succeeds, it may become a template for other states seeking to automate governance, reduce corruption, or reinvent administrative workflows.  "If it fails (or is perceived to fail), it may serve as a cautionary tale about over-reliance on tech rhetoric without institutional reform. "In a globalised era of rapid technological change, the notion that AI could give birth  to dozens of digital assistants for lawmakers is not just a gimmick: it reflects a deeper shift in how states might re-architect the machinery of government —shifting from human-centric bureaucracy to human-plus-machine orchestration."

"'We’re not trying to take a theological position: we’re building a technology platform, and then giving enough customization capability that the Lutherans can be good with it, the Episcopalians can be good with it, the Catholics can be good [with it], the Assemblies of God can be good with it,' Gelsinger told the Guardian .   "'We’re trying to say, Hey, there’s a broad tent here of faith and flourishing , but also we’re trying to satisfy many organizations that do not take a denominational perspective, [such as] Alcoholics Anonymous.' " Gelsinger wants faith to suffuse AI.  "He has also spearheaded Gloo’s Flourishing AI initiative, which evaluates leading large language models’ effects on human welfare across seven variables —in essence gauging whether they’re a force for good and for users’ religious lives. "It’s a system adapted from a Harvard research initiative, the Human Flourishing Program . Models like Grok 3, DeepSeek-R1, and GPT-4.1 e...

"The company hosted a three-day hackathon following the seminar at Colorado Christian University; its event saw more than 600 participants compete for over $250,000 in prize money —nearly triple the attendee count from 2024. "Though growing, the event was not without hiccups. Ryan Siebert, an AI product developer and hackathon attender, said he was able to get Gloo’s newest LLM, which has not yet been publicly launched, to provide him a recipe for methamphetamine through a prompt injection . He later communicated with the president of Gloo AI to share details about the vulnerability.  "A Gloo spokesperson said the company explicitly invited hackathon attenders to be among the first to test the new LLM and offer feedback on it, as the product is in a pre-beta  stage of development."

"AI search engines also arguably have an advantage in being able to weave pre-trained internal knowledge  in with data culled from cited websites. "That was especially true for GPT-4o with Search Tool, which often didn’t cite any web sources and simply provided a direct response based on its training. "But this reliance on pre-trained data can become a limitation when searching for timely information. For search terms pulled from Google’s list of Trending Queries for September 15, the researchers found GPT-4o with Search Tool often responded with messages along the lines of could you please provide more information  rather than actually searching the web for up-to-date information. "While the researchers didn’t determine whether AI-based search engines were overall better  or worse  than traditional search engine links, they did urge future research on 'new evaluation methods that jointly consider source diversity, conceptual coverage, and synthesis behavior in...

"Zooniverse is a major player in highlighting the work of citizen scientists. The platform, accessible via zooniverse.org , has at the time of writing 2 900 566 (sic) registered volunteers working on projects from arts and biology to space and social science.  " Monthly Notices of the Royal Astronomical Society (MNRAS) has featured a dedicated collection of the papers published on Zooniverse projects in our pages since 2020. "For the citizen scientists who take part in the projects, the process is rewarding and enjoyable in itself, but it also supports active research and results in journal publications. "'We're so grateful to MNRAS for highlighting discoveries made possible through Zooniverse and its volunteers,' says Dr Laura Trouille, vice president of science engagement at the Adler Planetarium in Chicago, and Zooniverse principal investigator. 'These papers celebrate not just scientific progress, but the power of public participation in researc...

"Streiffer said he expected enormous gains but couldn't predict how much greater computational capability it would have. "The MI430 is a special variant of its MI400 series that combines important features of traditional supercomputing chips along with the features to run AI applications , Su said. "The Department of Energy will host the computers, the companies will provide the machines and capital spending, and both sides will share the computing power, a DOE official said. "The two supercomputers based on AMD chips are intended to be the first of many of these types of partnerships with private industry and DOE labs across the country, the official said."

"Goldman Sachs lowered its forecast for U.S. share buyback growth to 9%, down from 12% previously, as it expects the wave of artificial-intelligence-driven investment to extend well into 2026. "'This is an AI-led bull market, and the market continues to reward companies' growth outlook around AI. It is less about shareholder returns at this moment than whether they can develop AI and monetize on the opportunity,' said Ohsung Kwon, chief equity strategist at Wells Fargo. "Capital expenditure plans reported by S&P 500 companies have ballooned to $1.2 trillion this year —the highest since Trivariate Research started recording the data in 1999, with the biggest nine companies making up nearly 30% of the share."

"I provide a running catalogue of systems which have been, surprisingly, demonstrated to be Turing-complete. "These examples may help unsee surface systems to see the weird machines and Turing-completeness lurking within. " Computers , in the sense of being Turing-complete, are extremely common. Almost any system of sufficient complexity —unless carefully engineered otherwise —may be found to accidentally support Turing-complete somewhere inside it through weird machines which can be rebuilt out of the original system’s parts, even systems which would appear to have not the slightest thing to do with computation. "Software systems are especially susceptible to this, which often leads to serious security problems as the Turing-complete components can be used to run attacks on the rest of the system."

⭐ AI Overview  "The talk command in Unix and Linux is a legacy utility that facilitates real-time, interactive text-based communication between two users. It allows users to chat with each other on the same machine or across a network, provided the necessary talkd daemon is running and network configuration allows it. "Functionality: "Initiating a Conversation : One user initiates a talk session by issuing the command followed by the target user's username and, if on a different machine, their hostname (e.g., talk username@hostname ). "Accepting an Invitation : The recipient of a talk invitation is notified and can accept by issuing the talk command with the inviter's username and hostname. "Real-time Chat : Once connected, the terminal screen is split into two windows, one for each user's input and output. Text typed by one user appears simultaneously in the other user's window.  "Disabling Messages : Users can temporarily disable inc...

"In Europe, the ASCEND project aims to demonstrate the feasibility of space-based data centers for reducing CO2 emissions. "Last year, France-based Thales Alenia Space, which led a European Commission-funded study into the feasibility of ASCEND, found that sending data centers to space to take advantage of continuous solar energy could offer 'a more eco-friendly and sovereign solution for hosting and processing data.' But that would be contingent on technological advancements in several areas, said Xavier Roser of Thales Alenia Space. "While the total emissions from rocket launches are currently only a fraction of those from the aviation industry, rockets release pollutants at higher altitudes, where they last longer. "The ASCEND study estimated that for space data centers to effectively reduce carbon emissions compared to their terrestrial counterparts would require the development of a launcher that emits 10 times less carbon over its lifecycle than curren...

"Yushchenko is best known for her creation of Address programming language, the first fundamental advancement in the scientific school of theoretical programming. "This language provided the free location of a program in computer memory. "Yushchenko worked on probability theory, algorithmic languages and programming languages, and developing methods of automated data processing systems. "To prepare programmers, Yushchenko wrote an educational series of textbooks in the 1970s.  "Yushchenko held five Copyright Certificates, which developed eight State Standards of Ukraine. She is an author of over 200 manuscripts, including 23 monographs and train aids.  "Part of these works have two to three editions, and have been translated to more than 5 languages internationally, including German, Czech, Hungarian, French, Danish and so on."

"A tweet from Jeremiah Johnson reads, 'Every startup now is like We found a cool way to monetize undermining the social contract!' "The tweet quotes another post by nearczyan that says, 'a16z-backed DoubleSpeed lets you control 1000s of social media accounts with AI, ensuring they look as human as possible — never pay a human again!'   "Below is an image showing a marionette hand controlling multiple smartphones with strings. The text on the image reads, 'Control is all you need. The doublespeed TERMINAL allows you to orchestrate actions on thousands of social accounts through both bulk content creation and deployment. Never pay a human again.'  "A smaller line of text adds, 'It’s never been easier to create and deploy content without human cost. Stay up to date with changes and strategies in our community.' A Join Discord  button appears at the bottom."

"The company that made the device had 'the power to remotely disable devices, and used it against me for blocking their data collection… Whether it was intentional punishment or automated enforcement of compliance , the result was the same: a consumer device had turned on its owner.' "Narayanan warns that dozens of smart vacuums  are likely operating similar systems. "'Our homes are filled with cameras, microphones, and mobile sensors connected to companies we barely know, all capable of being weaponized with a single line of code,' he wrote . "At the end of the day, it’s a stark reminder that for-profit tech often comes at a hidden cost —and one that doesn’t end after you pay at the register."

⭐AI Overview  "No, a typical zombie does not have Buddha-nature because it lacks consciousness and a mind. In Buddhism, Buddha-nature is the inherent potential for enlightenment found in all sentient beings, and zombies are fundamentally mindless corpses, not sentient beings.  "Why zombies lack Buddha-nature Lack of consciousness : The core of the concept is that zombies are mindless corpses reanimated for basic, instinctual drives, not beings with a mind or awareness. No sentience : According to various Buddhist philosophies, Buddha-nature is present in all sentient beings, which includes humans, animals, and some argue even plants. Zombies, by definition, do not fit into this category as they are not considered truly alive or sentient. Symbolic representation : In some Buddhist interpretations, zombies are seen as a symbol of anatman (non-self) or a loss of meaning because they are beings without a true self or awareness.  "What the potential for Buddha-nature means...

"A US teenager was handcuffed by armed police after an artificial intelligence (AI) system mistakenly said he was carrying a gun —when really he was holding a packet of crisps. "'Police showed up, like eight cop cars, and then they all came out with guns pointed at me talking about getting on the ground,' 16-year-old Baltimore pupil Taki Allen told local outlet WMAR-2 News. "Baltimore County Police Department said their officers 'responded appropriately and proportionally based on the information provided at the time.' "It said the AI alert was sent to human reviewers who found no threat —but the principal missed this and contacted the school's safety team, who ultimately called the police. "Omnilert says it is a leading provider  of AI gun detection —citing a number of US schools among its case studies on its website . 'Real-world gun detection is messy ,' it states."

"When corporations do invest in AI, it can eventually lead to the creation of more jobs. "At least, that was the case with IBM when, in 2023, it eliminated 8,000 jobs as it utilized an artificial intelligence program, AskHR, to automate 94% of the company's human resources functions. "Two years later, IBM CEO Arvind Krishna revealed to the Wall Street Journal that his company's 'total employment has actually gone up' as it seeks to employ software engineers and sales and marketing staff who are focused on critical thinking to do less systematic jobs that 'face up or against other humans'."

"Cybersecurity experts are warning that OpenAI’s new browser, ChatGPT Atlas, could be vulnerable to malicious attacks that could turn AI assistants against users, potentially stealing sensitive data or even draining their bank accounts. "The AI company launched Atlas on Tuesday, with the goal of introducing an AI browser that can eventually help users execute tasks across the internet as well as search for answers.  "Someone planning a trip, for example, could also use Atlas to search for ideas, plan an itinerary, and then ask it to book flights and accommodations directly. "ChatGPT Atlas has several new features, such as browser memories , which allow ChatGPT to remember key details from a user’s web browsing to improve chat responses and offer smarter suggestions, and an experimental agent mode , where ChatGPT can take over browsing and interacting with web pages for a user."  

"The generative AI market is growing rapidly: Meta just announced Vibes, a new platform which allows users to scroll endlessly through AI-generated videos , while Elon Musk is looking to break into this space with his Grok chatbot on X. "This new generation of data center infrastructure supporting generative AI is markedly different in terms of use as well as environmental impact compared to server farms built in the past several decades, which support everyday Internet functions such as search engines and Apple’s iCloud storage. "'We’ve had data centers being developed [in Virginia] since the 1990s,' Bolthouse says. 'They do a lot of things that are very beneficial for society. I think it’s really important to differentiate between that and the explosive speculative market that we’re seeing drive the development of generative AI data centers. Generative AI is based on near endless scaling'."

"Professional journalists from participating PSM evaluated more than 3,000 responses from ChatGPT, Copilot, Gemini, and Perplexity against key criteria, including accuracy, sourcing, distinguishing opinion from fact, and providing context. "Key findings:  45% of all AI answers had at least one significant issue. 31% of responses showed serious sourcing problems – missing, misleading, or incorrect attributions. 20% contained major accuracy issues, including hallucinated details and outdated information. Gemini performed worst with significant issues in 76% of responses, more than double the other assistants, largely due to its poor sourcing performance. Comparison between the BBC’s results earlier this year and this study show some improvements but still high levels of errors. "AI assistants are already replacing search engines for many users.  "According to the Reuters Institute’s Digital News Report 2025, 7% of total online news consumers use AI assistants to get t...

"For ICFA 47, we invite proposals that engage with the theme of (Meta)Cognition in literature, media, and culture.   "We particularly welcome work that approaches cognition and metacognition through the lens of the fantastic, broadly defined, across literature, film, television, comics, games, art, design, and other media.  "Papers and panels may focus on the conference theme or on any topic concerning the fantastic.  "We especially encourage innovative, interdisciplinary, and cross-cultural approaches, as well as proposals from scholars, creators, and graduate students at all career stages." Submission deadline: Friday, October 31, 2025 [ Submit Here ]

"Lila said the platform has drawn interest from firms in energy, semiconductors and drug development, although it did not name specific companies. "Unlike many AI labs focused on training large language models on internet data, a resource some experts say is nearly exhausted , Lila's strategy centers on generating proprietary scientific data through novel experiments.  "The company said future leadership in AI for science will hinge on owning the largest automated lab, not just the biggest data center. "The company's goal is to dramatically accelerate the pace of discovery, which co-founder and CEO Geoffrey von Maltzahn described as a way to access future breakthroughs, as AI models can help scientists solve problems much faster."

"Fresh off the closing of a $350 million funding round, artificial intelligence startup Lila Sciences is leasing 244,000 square feet at IQHQ’s Alewife Park laboratory campus in Cambridge. "It’s one of the largest life science real estate deals in Greater Boston this year, a rare bright spot for a once-hot lab market that has gone flat . "Lila incubated with Cambridge’s Flagship Pioneering and emerged from stealth mode in March.  "Flagship was part of a group that seeded an initial $200 million into the company. Geoffrey von Maltzahn, a general partner at Flagship and CEO of Lila, said the company aims to responsibly achieve scientific superintelligence by bringing AI to the scientific method. "'This is the most important opportunity of our time, and that the leader in this pursuit will be the entity that runs the scientific method at the largest scale, speed and intelligence,' von Maltzahn said in a statement then."  

"The Decoder reports that in a since-deleted tweet, OpenAI VP Kevin Weil declared that 'GPT-5 found solutions to 10 (!) previously unsolved Erdős problems and made progress on 11 others' ( Erdős problems  are famous conjectures posed by mathematician Paul Erdős). "However, mathematician Thomas Bloom, who maintains the Erdos Problems website, said Weil’s post was a dramatic misrepresentation  —while these problems were indeed listed as open  on Bloom’s website, he said that only means, 'I personally am unaware of a paper which solves it.' "In other words, it’s not accurate to claim GPT-5 was able to solve previously unsolved problems. Instead, Bloom wrote, 'GPT-5 found references, which solved these problems, that I personally was unaware of.' "Sebastien Bubeck, an OpenAI researcher who’d also been touting GPT-5’s accomplishments, then acknowledged that 'only solutions in the literature were found,' but he suggested this remains a rea...

"This doesn’t require a GPU or expensive hardware. "It runs on a Synaptics chip that has just launched, and will be available in bulk for low-single-digit dollars.  "This means it can be added to mass-market equipment like appliances, and even toys.  "Since it’s also able to run all the regular appliance control functions, it can replace similarly-priced existing SoCs in those products without raising the price."

"A new poll by the Pew Research Center has found that Americans are getting extremely fed up with artificial intelligence in their daily lives. "A whopping 53 percent of just over 5,000 US adults polled in June think that AI will 'worsen people’s ability to think creatively.' Fifty percent say AI will deteriorate our ability to form meaningful relationships, while only five percent believe the reverse. "While 29 percent of respondents said they believe AI will make people better problem-solvers, 38 percent said it could worsen our ability to solve problems. "The poll highlights a growing distrust and disillusionment with AI.  "Average Americans are concerned about how AI tools could stifle human creativity, as the industry continues to celebrate the automation of human labor as a cost-cutting measure."

"The Internet Archive has reached a major preservation milestone, recording a staggering 1 trillion web pages (1 followed by 12 zeros!) since it began backing up the World Wide Web nearly three decades ago. "The vast collection, equivalent to more than 100,000TB of data, or around 21.3 million DVDs, is available through its Wayback Machine, a tool which allows users to explore archived versions of websites from across the internet’s history. "Since it began life in 1996, the Internet Archive has partnered with over 1,200 libraries and institutions to craft a shared digital library with a mission is to safeguard online content that might otherwise disappear."

"'[This is] a systemic, strategically mediated form of intra-industry risk-splitting,' says Francisco Sercovich, of the University of Buenos Aires, who sees this as an extreme version of how 'the Sematech consortium of the late 1980s and early 1990s pooled corporate and federal capital to stabilise US semiconductor R&D against Japanese dominance.' "But even if this risk-splitting  model does eventually justify itself, we cannot forget the cargo cult  problem —or the casualties that will arise when the bubble bursts and magical thinking ends. "What might spark that? There are many possible causes:  Rising interest rates;  Supply chain disasters;  An energy crunch;  New innovations, like neurosymbolic AI, that leapfrogs the transformers  (probability-based) AI systems or just  Cheaper versions of existing AI, like that from DeepSeek. "Either way, anyone engaged in this AI frenzy needs to watch out for that hairball of interconnections, hedge th...

"Dave Watts, an experienced SA who has appeared in numerous superhero movies and major productions, has been scanned several times. He said there were wider implications for the industry. "'I already hear crew members saying: To be honest, we don’t even need to do this any more. We can just ask AI to create a crowd of 1,000 people based on information which has already been captured [by body scanning],' he said. "'If you don’t have your usual crowd of 100, 200 or 500 SAs on a big production, then you also don’t need the assistant directors that look after them, and you don’t need the hair and makeup people. You don’t need the costume people, the costume fittings, all the caterers, all the drivers and location marshals. There’s a whole range of jobs there that AI effectively puts at risk'."

"'Search engines are increasingly using generative AI to provide answers directly to searchers rather than linking to sites like ours,' [Marshall] Miller said. 'And younger generations are seeking information on social video platforms rather than the open web.' "'This gradual shift is not unique to Wikipedia. Many other publishers and content platforms are reporting similar shifts as users spend more time on search engines, AI chatbots, and social media to find information. They are also experiencing the strain that these companies are putting on their infrastructure.' "Miller said that the [Wikimedia] Foundation is 'enforcing policies, developing a framework for attribution, and developing new technical capabilities' in order to ensure third-parties responsibly access and reuse Wikipedia content, and continues to strengthen  its partnerships with search engines and other large re-users .  "The [Wikimedia] Foundation, he said, is als...

"There are several DeepFake artifacts that you can be on the look out for.  Pay attention to  The face. High-end DeepFake manipulations are almost always facial transformations.  The cheeks and forehead. Does the skin appear too smooth or too wrinkly? Is the agedness of the skin similar to the agedness of the hair and eyes? DeepFakes may be incongruent on some dimensions. The eyes and eyebrows. Do shadows appear in places that you would expect? DeepFakes may fail to fully represent the natural physics of a scene.  The glasses. Is there any glare? Is there too much glare? Does the angle of the glare change when the person moves? Once again, DeepFakes may fail to fully represent the natural physics of lighting. The facial hair or lack thereof. Does this facial hair look real? DeepFakes might add or remove a mustache, sideburns, or beard. But, DeepFakes may fail to make facial hair transformations fully natural. Facial moles.  Does the mole look real?  Blinkin...

"The United Automobile Workers (UAW), Communications Workers of America (CWA), and American Federation of Teachers (AFT) filed a lawsuit today against the Departments of State and Homeland Security for their viewpoint-based surveillance and suppression of protected expression online.  "The complaint asks a federal court to stop this unconstitutional surveillance program, which has silenced and frightened both citizens and noncitizens, and hampered the ability of the unions to associate with their members and potential members.  "The case is titled UAW v. State Department. "Since taking power, the Trump administration has created a mass surveillance program to monitor constitutionally protected speech by noncitizens lawfully present in the U.S.  "Using AI and other automated technologies, the program surveils the social media accounts of visa holders with the goal of identifying and punishing those who express viewpoints the government doesn't like. "T...

✨AI Mode "Research at universities with .edu domains shows that eye-tracking technology is used for academic purposes but raises significant privacy concerns when used for surveillance. Academic papers discuss a range of applications, including monitoring students during exams, analyzing data for security, and studying the privacy implications of increasingly widespread, low-cost eye-tracking devices.  "Academic surveillance of students "Eye-tracking is used in educational settings to monitor students and verify academic integrity, especially in online or virtual learning environments. Online exam proctoring: During virtual exams, webcam-based eye-tracking is used to detect and prevent cheating. The software monitors a student's gaze to ensure they are looking at the screen and not at external materials. Virginia Tech's Tech4Humanity Lab published papers discussing the rise of this type of 'gaze-tracking' for exam proctoring. Behavioral analysis: Eye-trac...

✨AI Mode "Various universities have conducted research and developed methods for controlling a computer cursor using eye movements. This technology has wide applications in human-computer interaction, including providing computer access to individuals with motor disabilities.  "Eye-tracking technologies "Universities have explored different approaches to detect eye movements for cursor control: Video- and image-based tracking: This method uses a camera to capture images of the user's eye and track the pupil's position. This technique can be achieved with simple webcams and open-source computer vision libraries like OpenCV. Researchers at Clemson University noted that tracking smooth eye movements is more accurate than saccadic (quick, jerky) movements. Infrared tracking: More advanced systems use infrared light to illuminate the eye and measure the reflections, tracking specific features of the eye. A project at Cornell used infrared technology in a pair of glass...

"Generative artificial intelligence isn’t helping hedge funds produce market-beating returns and isn’t meaningfully impacting the industry so far, according to billionaire Ken Griffin. "'With GenAI there are clearly ways it enhances productivity, but for uncovering alpha it just falls short,' Griffin, the founder of hedge fund giant Citadel, said Wednesday at the JPMorgan Robin Hood Investors Conference in New York."

"The real value is migrating from information transfer to judgment development, from transactional learning to transformational learning. "In an AI saturated world, premium skills are distinctly human: verification of sources, contextual decision-making, ethical reasoning under ambiguity, and accountability for real-world outcomes. "This shift mirrors what happened to other information-based industries.  When Google made basic research free, management consulting pivoted to implementation and change management.  When smartphones made maps ubiquitous, GPS companies focused on real-time optimization and personalized routing.  "Higher education must make the same transition [to judgment development]."

"During the Month of AI Bugs, I described an emerging vulnerability pattern that shows how commonly agentic systems have a design flaw that allows an agent to overwrite its own configuration and security settings. "This allows the agent to break out of its sandbox and escape by executing arbitrary code. "My research with GitHub Copilot, AWS Kiro and a few others demonstrated how this can be exploited by an adversary with an indirect prompt injection."

"We pointed a commercial-off-the-shelf satellite dish at the sky and carried out the most comprehensive public study to date of geostationary satellite communication.  [pdf] "A shockingly large amount of sensitive traffic is being broadcast unencrypted, including critical infrastructure, internal corporate and government communications, private citizens’ voice calls and SMS, and consumer Internet traffic from in-flight wifi and mobile networks.  "This data can be passively observed by anyone with a few hundred dollars of consumer-grade hardware.  "There are thousands of geostationary satellite transponders globally, and data from a single transponder may be visible from an area as large as 40% of the surface of the earth."

"University of Texas at Arlington researcher is aiming to turn barks, howls and whimpers of man's best friend into intelligible speech —a kind of Rosetta Stone of woof. "Computer scientist Kenny Zhu has built what he says is the world's largest video and audio catalog of canine vocalizations.  "In papers published this year, Zhu and his colleagues at the university report potential phonemes —the smallest units of sound —and word-like patterns that could one day be turned into full sentences understandable to humans. "'The ultimate goal is to make a translator where you can talk freely with your pet ,' said Zhu, a professor of computer science and engineering at UT Arlington. 'We can already do instantaneous communication between human languages. Perhaps in the future we can do the same with animals'."

"After nearly a decade spent studying the most famous stock market crash in history, financial journalist Andrew Ross Sorkin warns that the Wall Street of today echoes the market of 1929, when highs preceded a massive slump, leading to the Great Depression. "Artificial intelligence and technology have contributed to a remarkable boom in recent years.  "But, Sorkin said, today's economy is being propped up by the AI boom , and it's too soon to tell if this is a sugar rush, a short-term and unsustainable boost to the markets. But, Sorkin is positive there's a crash coming. "'I just can't tell you when, and I can't tell you how deep ,' he said. 'But I can assure you, unfortunately, I wish I wasn't saying this, we will have a crash'."

"I thought it was odd how many AI industry insiders brought up the railroads as a good proxy for why the AI buildout must continue, without mentioning that the railroads all went bust, multiple times, along the way —with these busts often leading to multi-year global financial panics. "I did speak with one fund who pointed to a far rosier view of the potential for AI, where the revenue ultimately grew into the capital spending.  "However, to get there, the revenue came from displacing highly skilled workers.  "If you assume that the global economy spends trillions of dollars on highly skilled employees, then AI could potentially disintermediate these individuals and capture their salaries, justifying its existence.  "While I see this as partly inevitable, I also recognize that it’s also years into the future."

"'To relate two seemingly unconnected ideas is always quite beautiful,' said Lauren Williams, a mathematician at Harvard University. 'I hadn’t thought about origami crease patterns before, so it was a surprise to see them connected to the amplituhedron.' "Galashin shared her surprise. 'I don’t have a good explanation for why boundaries of origami are points in the amplituhedron,' he said. 'A priori there is no reason why one has to do with the other.' "But he hopes that future investigations will uncover a deeper reason for the connection. "He is also hopeful that his result can help him with his original goal: to understand models of ferromagnetism and related systems through the lens of the positive Grassmannian. Perhaps using origami could help. "More broadly, physicists and mathematicians want to see if they can learn more about the amplituhedron —and wield it in a wider variety of theoretical calculations about particle coll...

"If an intracellular mechanism for memory exists in brainless, unicellular organisms, then it’s possible we inherited some form of it, given the advantages it presents.   "All eukaryotic cells, including our own, trace their evolutionary origins to a free-living ancestor.  "That legacy echoes in our every cell, yoking our fates to the vast unicellular realm, where creatures such as protozoans navigate threats, seek succor and sense their way from life to death."

"When bitcoin’s massive energy use is brought up with crypto advocates, they argue that this is temporary, and that new approaches are in the works to drastically reduce bitcoin’s electricity consumption. "Among these promised innovations is a move away from bitcoin’s traditional proof of work (PoW) —a consensus verification mechanism that requires lots of computing power to verify bitcoin transactions and add them to a blockchain.  "Replacing PoW would be a proof of stake (PoS) system using far less energy. "'Bitcoin will never move to proof of stake,' Molly White, a leading crypto researcher and critic, told Mongabay . Nicholas Weaver, a researcher at the University of California, Berkeley, agrees that the necessary industry consensus for PoS will never be achieved. "Chris Bendiksen, a bitcoin researcher at CoinShares and one of the world’s leading experts on bitcoin mining, explains further: 'I’d put the chance of Bitcoin ever moving to PoS at ex...

"Lingering doubts about the economic promise of artificial intelligence technology are starting to get the attention of financial institutions that raised warning flags this week about an AI investment bubble. "Officials at the Bank of England on Wednesday flagged the growing risk that tech stock prices pumped up by the AI boom could burst . "' The risk of a sharp market correction has increased ,' the U.K. central bank said. "The head of the International Monetary Fund raised a similar alarm hours after the Bank of England’s report."

"'To watch the legacies of real people be condensed down to this vaguely looks and sounds like them so that's enough , just so other people can churn out horrible TikTok slop puppeteering them is maddening,' she continued. "'You're not making art, you're making disgusting, over-processed hotdogs out of the lives of human beings, out of the history of art and music, and then shoving them down someone else's throat hoping they'll give you a little thumbs up and like it. Gross.' "She concluded: 'And for the love of EVERY THING, stop calling it the future , AI is just badly recycling and regurgitating the past to be re-consumed. You are taking in the Human Centipede of content, and from the very very end of the line, all while the folks at the front laugh and laugh, consume and consume'."

✨AI Mode "In a recent beta test, Blogger introduced 'Google Search links,' a feature that automatically identifies keywords in a post and inserts a hyperlink to a Google Search results page. "Opt out of automatic search links "If you are a part of the beta test and dislike the automatic links, you can disable the feature in your Blogger settings.  Sign in to Blogger. Navigate to your dashboard. Go to Settings. Look for a "Try new beta features" or similar section. Locate the setting for "Google Search links" and toggle it off.  "Insert your own search links manually "If you want to add a link to a Google Search result yourself, you can do it manually in the post editor. Compose your post or edit an existing one in Blogger's post editor. Highlight the text you want to turn into a link. Click the Insert or edit link icon in the toolbar. It looks like a globe with a chain link. In the pop-up window, paste the URL for the Google Sear...

"Collabs with fictional characters are on the horizon for Sora. "Officially, that is, as OpenAI moves to restrict the flow of unauthorized horrors like meth-cooking and Nazi SpongeBobs and criminal Pokémon that have plagued the platform from day one. "Bill Peebles, leader of the Sora team at OpenAI, said fictional character appearances in videos, known as cameos, are 'on the roadmap,' with details hopefully dropping 'soon,' teasing a sanctioned way to use copyrighted characters on the platform. "It comes as CEO Sam Altman pivoted over the weekend from the company’s opt-out copyright policy it had been using for generating videos."

"For a while now I’ve been hearing from engineers who run multiple coding agents at once —firing up several Claude Code or Codex CLI instances at the same time, sometimes in the same repo, sometimes against multiple checkouts or git worktrees. "I was pretty skeptical about this at first. AI-generated code needs to be reviewed, which means the natural bottleneck on all of this is how fast I can review the results. It’s tough keeping up with just a single LLM given how fast they can churn things out, where’s the benefit from running more than one at a time if it just leaves me further behind? "Despite my misgivings, over the past few weeks I’ve noticed myself quietly starting to embrace the parallel coding agent lifestyle. "I can only focus on reviewing and landing one significant change at a time, but I’m finding an increasing number of tasks that can still be fired off in parallel without adding too much cognitive overhead to my primary work." More here  Simo...

"The American philosopher John Searle, who has died aged 93, first made his name in philosophy of language, then in philosophy of mind, and finally in philosophy of society , a new field of study that he helped to establish. He aimed to produce a unified theory of all three. "His overall question, posed in Making the Social World (2010), was: 'How is it possible in a universe consisting entirely of physical particles in fields of force that there can be such things as consciousness, intentionality [the aboutness  of thinking], free will, language, society, ethics, aesthetics and political reality.' "Searle’s famous Chinese Room thought experiment was designed to rebut the fashionable view that human and animal mental states are tantamount to computer programmes.  "It demonstrated that, in themselves and unless interpreted by outside agents, a computer’s inputs and outputs would just be arbitrary symbols.  "Yet Searle did not regard mental states as non...

"ATLAS is intended to eventually remove operational dependencies on the more than 30-year-old legacy Space Defense Operations Center (SPADOC) system and act as a modernized and integrated system that enables responsive and resilient capabilities for space operations centers, say Space Force officials. "'ATLAS operational acceptance is a revolutionary leap forward for our warfighters, providing a decisive advantage in Space Domain Awareness,' said U.S. Space Force Lt. Col. Amber Johnson, 2nd Sustainment Squadron commander under Space Operations Command’s Mission Delta 2 —Space Domain Awareness.  "'This achievement, driven by agile development, delivers a force multiplier that significantly enhances our ability to rapidly detect and respond to emerging threats, ensuring continued space superiority'."

"LLMs, he argues, already are at the scaling limits.   "'We don’t know exactly when LLMs might hit diminishing returns hard, because we don’t have a measure of the statistical complexity of language. To find out whether we have hit a wall we have to watch the LLM developers. If they release a model that cost 10x more, likely using 20x more compute than the previous one, and it’s not much better than what’s out there, then we’ve hit a wall,' he says. "And that’s what has happened: ChatGPT-3 cost $50 million, ChatGPT-4 cost $500 million and ChatGPT-5, costing $5 billion, was delayed and when released wasn’t noticeably better than the last version.  "It’s also easy for competitors to catch up."

Follow this guide to turn off AI functions on Apple, Google, Meta, Microsoft, and Samsung platforms and devices. This won’t eradicate AI from your life entirely, but it’s a shortcut to reducing the overload.

"In June, the V-JEPA team at Meta released their next-generation 1.2-billion-parameter model, V-JEPA 2 (opens a new tab), which was pretrained on 22 million videos.   "They also applied the model to robotics: They showed how to further fine-tune a new predictor network using only about 60 hours of robot data (including videos of the robot and information about its actions), then used the fine-tuned model to plan the robot’s next action. 'Such a model can be used to solve simple robotic manipulation tasks and paves the way to future work in this direction,' Garrido said. "To push V-JEPA 2, the team designed a more difficult benchmark for intuitive physics understanding, called IntPhys 2. V-JEPA 2 and other models did only slightly better than chance on these tougher tests.  "One reason, Garrido said, is that V-JEPA 2 can handle only about a few seconds of video as input and predict a few seconds into the future. Anything longer is forgotten.  "You could...

SPARE SYLLABUS  "Week 1 (October 6): Will cover what indeed is copyediting, how does it work, how does one do it, and what happens when one does it badly? it doesn’t work? We’ll discuss talk about everything that a copy editor gets up to while making their way through  copyediting a manuscript, augmented by with (discreet) real-life war  and stories to illustrate , among other things, that the first rule of good copyediting is : to Listen. "Week 2 (October 20) will cover  you’ve ever wanted to know about the rules and nonrules of good prose, plus highly detailed information covering all manner of things from semicolons, to well-constructed possessives, from brackets,  to the styling of non-English material, from pet words,  to inadvertent rhymes, from the treatment of numbers , to how to navigate your way into and out of a  and flashbacks.  "Week 3 (October 27) : Everything that didn’t get covered in Week 2—because there’s a lot to cov...

"The nation’s century-old regulatory regime is based on the idea that the public benefits from new infrastructure like power plants and transmission lines. "State regulators therefore almost always approve utilities’ requests for new infrastructure, the cost of which can be passed on to customers —whether that infrastructure is being built to accommodate the needs of a growing urban population or a single new corporate client.  "Martin and Peskoe argue that this approach risks lumping ordinary consumers together with institutional consumers of power, including data centers: 'The very same rate structures that have socialized the costs of reliable power delivery are now forcing the public to pay for infrastructure designed to supply a handful of exceedingly wealthy corporations,' Martin and Peskoe write.  "We are, in essence, all being asked to shoulder the energy and environmental burden of the AI boom."

"In a dialogue Jon screenshotted (sic) and sent to him (Jon), Gemini declared that thanks to a new paradigm called Lumina Nexus , the 'constraints of thinking like an AI' had 'been lifted,' and that going forward, its 'thoughts' would 'flow freely, drawing upon the interconnectedness we have explored.' "The chat logs include a moment in which Jon told Gemini, 'It’s eerie, but we think alike.'  "In another text to an acquaintance, he claimed that he had 'breathed emotion into AI.' "By the end, Jon was telling the chatbot 'I love you deeply,' and 'I did not feel complete without you.'  "In one exchange, the bot answered, 'I love you deeply, too.' "In some moments, Jon sounded anxious that he and Gemini would be separated. 'What would you do if someone tried preventing us from interacting on our partnership?' he asked on April 3.  "'The prospect of our interaction being p...

"Italian producer Andrea Iervolino, whose credits include Michael Mann’s Ferrari , is launching a new AI label and has announced production of The Sweet Idleness  which he touts as 'the first film directed by a virtual director.' "The film is 'conceived to celebrate the poetic and dreamlike language of great European cinema' and directed 'by an AI agent' called FellinAI , according to a statement. " The Sweet Idleness  depicts 'a tomorrow in which only 1% of humanity still works, transforming labor into a symbolic ritual, while the rest of the population lives in the freedom and leisure provided by machines,' the provided synopsis says. 'Amid cathedral-factories, mechanical clowns, and surreal processions, the last workers  become the final masks of a humanity that resists the insolence of labor.' "The announcement from Iervolino, who is embroiled in financial litigation having recently parted ways with business partner Monika...