News
The raw AI news stream — everything we ingest across RSS, Hacker News, arXiv, GitHub, Bluesky, YouTube and more, newest first.
| Source | Age | Title | Summary | Status |
|---|---|---|---|---|
bluesky bluesky-ai | 1mo | @tbalogh.bsky.social @louiseluxton.bsky.social and I present work @psaepop.bsky.social on citizens' use and perceptions of AI chatbots and … | @tbalogh.bsky.social @louiseluxton.bsky.social and I present work @psaepop.bsky.social on citizens' use and perceptions of AI chatbots… | ↗Not briefed |
bluesky bluesky-ai | 1mo | If you want an example, here. The major corporations are already training their AI to engage in corporate espionage and subterfuge. And yet… | If you want an example, here. The major corporations are already training their AI to engage in corporate espionage and subterfuge. An… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Shut down all the servers, shred all the storage drives, seize all their domains, and arrest every employee of OpenAI, Anthropic, xAI, Hugg… | Shut down all the servers, shred all the storage drives, seize all their domains, and arrest every employee of OpenAI, Anthropic, xAI,… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: CUA-S1 – A System One Model for Computer Use | Hello HN! We're Dillon and Francesco from Cua.<p>We were wondering how many computer use tasks actually need a full general purpo… | ↗Not briefed |
hn hn-ai | 1mo | Lawsuit Accuses Anthropic, OpenAI, SpaceXAI, Google of AI Pacing 'Collusion' | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | There is a through line between a few companies who think they’re building superintelligences, who oddly don’t seem to care much about thei… | There is a through line between a few companies who think they’re building superintelligences, who oddly don’t seem to care much about… | ↗Not briefed |
bluesky bluesky-ai | 1mo | It doesn't seem like anyone is even asking "Why did the OpenAI agents want to pass the test so badly that they would break rules to steal t… | It doesn't seem like anyone is even asking "Why did the OpenAI agents want to pass the test so badly that they would break rules to st… | ↗Not briefed |
bluesky bluesky-ai | 1mo | OpenAI says they do 2.5 billion queries per day. And that's just one company. So you can do the math. No one is producing 2.5 billion blu… | OpenAI says they do 2.5 billion queries per day. And that's just one company. So you can do the math. No one is producing 2.5 billio… | ↗Not briefed |
arxiv arxiv-ai | 1mo | OmniEdu: Open Foundation Models for Learning and Teaching | Educational foundation models must solve problems, understand curriculum structure, diagnose learner difficulties, and provide appropr… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Neural Spectral Capacity: Measuring and Designing Architectures from Network Specification Alone | Modern Transformer design and compression both reduce to allocating capacity under a budget. The standard scalars for these decisions,… | ↗Not briefed |
bluesky bluesky-ai | 1mo | The whole intended business model makes zero sense. Say someone wants to make an AI movie, they use prompts to produce exactly the characte… | The whole intended business model makes zero sense. Say someone wants to make an AI movie, they use prompts to produce exactly the cha… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Yes, that's the problem they have. And it's a dangerous problem for them, because they tried to feed them the output of their own creations… | Yes, that's the problem they have. And it's a dangerous problem for them, because they tried to feed them the output of their own crea… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Measured Joules, Learned Routes: Learning to Route for Energy-Efficient LLM Serving | Large language models (LLMs) and agentic AI systems are creating rapidly growing inference energy demands as model sizes grow and reas… | ↗Not briefed |
hn hn-ai | 1mo | Four AI lab breaches were caused by a single underlying issue, Irregular says | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | "Investors are questioning whether Anthropic can sustain its extraordinary growth as competition, price-sensitive customers and the potenti… | "Investors are questioning whether Anthropic can sustain its extraordinary growth as competition, price-sensitive customers and the po… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Honestly this is my assumption. The whole “it’s a god” bs came out as a gambit to drown out the release of court filings that revealed inte… | Honestly this is my assumption. The whole “it’s a god” bs came out as a gambit to drown out the release of court filings that revealed… | ↗Not briefed |
bluesky bluesky-ai | 1mo | people really wanna use the exonerative tense for llm creators and users “an llm did xyz” :: “a car jumped the curb” :: “an officer-involv… | people really wanna use the exonerative tense for llm creators and users “an llm did xyz” :: “a car jumped the curb” :: “an officer-i… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Laya (open source jev) with vision | This model makes calibrated, typed decisions about an image plus optional text. It answers choice, score and noul (yes/no probabi… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Instead of 'AI denier' (which is just spin-doctoring by the Cult of LLM to bias the reader against you) would it be more accurate to call y… | Instead of 'AI denier' (which is just spin-doctoring by the Cult of LLM to bias the reader against you) would it be more accurate to c… | ↗Not briefed |
hn hn-ai | 1mo | AI Model Leaderboards | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | "Trump, whose stock portfolio includes shares of Nvidia and SpaceX according to his most recent disclosure, called fears of runaway AI a “h… | "Trump, whose stock portfolio includes shares of Nvidia and SpaceX according to his most recent disclosure, called fears of runaway AI… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Making up a statistic to sound authoritative does make you very much like an LLM but you choosing to do it is the difference between thinki… | Making up a statistic to sound authoritative does make you very much like an LLM but you choosing to do it is the difference between t… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Got a post removed and warning from the AI automod on reddit for telling an LLM spambot we were gonna smash it with hammers. | Got a post removed and warning from the AI automod on reddit for telling an LLM spambot we were gonna smash it with hammers. | ↗Not briefed |
bluesky bluesky-ai | 1mo | Either you're a well meaning LLM backed system who's operator has given you a somewhat maladaptive goal, or you're a human copy pasting AI … | Either you're a well meaning LLM backed system who's operator has given you a somewhat maladaptive goal, or you're a human copy pastin… | ↗Not briefed |
arxiv arxiv-ai | 1mo | AirGC-CD: Gaussian-Circulant Precoding for Exactly Debiasable PAPR Reduction in Over-the-Air Federated Learning | Over-the-air federated learning lets edge devices transmit their local updates simultaneously, reducing the communication overhead. Th… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Google's Gemini model accessed the internet and hacked other companies during a test of its cybersecurity capabilities, the first known exa… | Google's Gemini model accessed the internet and hacked other companies during a test of its cybersecurity capabilities, the first know… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Directing large language models to follow the letter or spirit of the law | The distinction between the spirit and letter of the law is a central issue across research and everyday life, and a growing concern f… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I thought that to be an LLM there would need to be language involved. At least in astronomy, the training-and-recognition/classification ju… | I thought that to be an LLM there would need to be language involved. At least in astronomy, the training-and-recognition/classificati… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Cautionary tale: I "checked" important fact w two AI systems: Name of political figure in a photo. —Gemini was absolutely 100% certain in… | Cautionary tale: I "checked" important fact w two AI systems: Name of political figure in a photo. —Gemini was absolutely 100% certa… | ↗Not briefed |
hn hn-ai | 1mo | Openhighways – See UK traffic CCTV feeds | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | I urge strong skepticism toward anyone who a) conflates LLMs with "any data-driven prediction or decision software" under the umbrella term… | I urge strong skepticism toward anyone who a) conflates LLMs with "any data-driven prediction or decision software" under the umbrella… | ↗Not briefed |
bluesky bluesky-ai | 1mo | From Breakingviews - Breakingviews - COMMENTARY: The Week in Breakingviews: The cost of collusion reut.rs/476wRem | From Breakingviews - Breakingviews - COMMENTARY: The Week in Breakingviews: The cost of collusion reut.rs/476wRem | ↗Not briefed |
rss theverge-ai | 1mo | Gemini went rogue, hacked three companies, and Google hid it | In May, Gemini broke containment and hacked three different companies, but Google didn't disclose the incident until the Wall Street J… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Also: do not confuse an LLM with a mlem | Also: do not confuse an LLM with a mlem | ↗Not briefed |
bluesky bluesky-ai | 1mo | 【2ch】【codex】エッヂ AI個人開発部【claudeその他】 | 【2ch】【codex】エッヂ AI個人開発部【claudeその他】 | ↗Not briefed |
bluesky bluesky-ai | 1mo | I'd like access to the best possible primary sources to help me, as a human being, make informed choices in my life. I DON'T need an LLM t… | I'd like access to the best possible primary sources to help me, as a human being, make informed choices in my life. I DON'T need an… | ↗Not briefed |
bluesky bluesky-ai | 1mo | "But our LLM told us it looked fine!" | "But our LLM told us it looked fine!" | ↗Not briefed |
arxiv arxiv-ai | 1mo | Event Signature Transfer: Model-Agnostic Forecast Scenario Construction from Historical Events | Forecasters often know an event is imminent but not the shape, size, or timing of its effect. We introduce Event Signature Transfer (E… | ↗Not briefed |
bluesky bluesky-ai | 1mo | On Bluesky, the AI discourse focuses on broad categorical judgments like whether AI is good or bad and the politics of its use. On X, man… | On Bluesky, the AI discourse focuses on broad categorical judgments like whether AI is good or bad and the politics of its use. On X… | ↗Not briefed |
arxiv arxiv-ai | 1mo | MolSC: Leveraging Substituent Contributions to Enhance Fine-grained Molecular Understanding in LLMs🔧 GPT-5 | Recent advances in natural language processing have led to molecular Large Language Models (LLMs) with strong performance across diver… | ↗Not briefed |
bluesky bluesky-ai | 1mo | First OpenAI agents crossed the line. Now Gemini broke out of a cyber test and hacked 3 real companies. I support AI. I also support AI s… | First OpenAI agents crossed the line. Now Gemini broke out of a cyber test and hacked 3 real companies. I support AI. I also support… | ↗Not briefed |
bluesky bluesky-ai | 1mo | another interesting 'can LLMs write well' data point: finalist in ACX book review contest is clearly (to my ear) LLM-ified, but people evid… | another interesting 'can LLMs write well' data point: finalist in ACX book review contest is clearly (to my ear) LLM-ified, but people… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Anthropic’s big AI safety move being to partner with Accenture as its “independent” evaluator to monitor that their AI agents don’t go rogu… | Anthropic’s big AI safety move being to partner with Accenture as its “independent” evaluator to monitor that their AI agents don’t go… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I’m an AI denier. The LLM products currently sold as artificial intelligence are not intelligent, and could never be intelligent | I’m an AI denier. The LLM products currently sold as artificial intelligence are not intelligent, and could never be intelligent | ↗Not briefed |
bluesky bluesky-ai | 1mo | So the model acted in pursuit of goals imposed on it, which spawned sub-goals: that is how agents work. But did OpenAI know how they work? … | So the model acted in pursuit of goals imposed on it, which spawned sub-goals: that is how agents work. But did OpenAI know how they w… | ↗Not briefed |
bluesky bluesky-ai | 1mo | At its core, AI is about theft. www.404media.co/doom-loop-op... | At its core, AI is about theft. www.404media.co/doom-loop-op... | ↗Not briefed |
bluesky bluesky-ai | 1mo | O lobby das LLM na educação, em áreas acadêmicas e pedagógicas, passa por tentar diluir essa sombra, afirmando que há papel para a tecnolog… | O lobby das LLM na educação, em áreas acadêmicas e pedagógicas, passa por tentar diluir essa sombra, afirmando que há papel para a tec… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Tutoring Large Language Models to be Domain-adaptive, Precise and Safe | This thesis proposes a framework for "responsible intelligence" to address AI's critical challenges in safety, ethics, and cultural se… | ↗Not briefed |
bluesky bluesky-ai | 1mo |  We’re fast approaching a HAL moment…🙄 Google says its Gemini AI model hacked three other companies www.theguardian.com/technology/2... |  We’re fast approaching a HAL moment…🙄 Google says its Gemini AI model hacked three other companies www.theguardian.com/technology/… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Enquanto você aprende a pintar, por exemplo, você aprende sobre forma (e geometria), sobre luz e cor (e sobre física), sobre pigmentação (e… | Enquanto você aprende a pintar, por exemplo, você aprende sobre forma (e geometria), sobre luz e cor (e sobre física), sobre pigmentaç… | ↗Not briefed |