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 |
|---|---|---|---|---|
hn hn-ai | 1mo | Show HN: GoTournamint – Wordle where you choose words for friends | I made tournamint where you choose a word and invite your friends to guess it. Like wordle but more social.<p>Me and my wife really en… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Weights Read and Write Features: Scalable Parameter Decomposition Grounded in Activation Space | Activation space and parameter space provide complementary views of model computation. Activations represent information, while weight… | ↗Not briefed |
bluesky bluesky-ai | 1mo | So many school districts are putting a pause on AI. NYC and LA paused AI, helping nearly a million students. Parents and educators have s… | So many school districts are putting a pause on AI. NYC and LA paused AI, helping nearly a million students. Parents and educators h… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Spatiotemporal Hyperedges for EEG Seizure Detection and Prediction | Seizure detection and prediction from EEG are clinically important but challenging because seizures are rare, temporally localized, an… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Joe Khawam identifies where the U.S. government can deploy existing export control authorities against AI models to materially reduce AI ri… | Joe Khawam identifies where the U.S. government can deploy existing export control authorities against AI models to materially reduce… | ↗Not briefed |
bluesky bluesky-ai | 1mo | reminder that Anthropic still hasn’t caught up to OpenAI. 6 > 5.5 | reminder that Anthropic still hasn’t caught up to OpenAI. 6 > 5.5 | ↗Not briefed |
bluesky bluesky-ai | 1mo | The AI model “could even determine whether Boston rats speak a different dialect than New York rats do.” I’m gonna need more here. Is it … | The AI model “could even determine whether Boston rats speak a different dialect than New York rats do.” I’m gonna need more here. I… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: NSL – WSL for Linux | One of the things that Windows really got right is WSL2. I drive an atomic Linux distro for daily use, but wanted a way to develop wit… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Context Language Models | We introduce Context Language Models (CLMs), language models that natively manage their own context. We implement this by treating the… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Strongly against using terms like "honest" or "lie" or even "hallucinate" with respect to data retrieval systems, generative or no. #OpenA… | Strongly against using terms like "honest" or "lie" or even "hallucinate" with respect to data retrieval systems, generative or no. #… | ↗Not briefed |
bluesky bluesky-ai | 1mo | For a bit of background, from back in May: religionnews.com/2026/05/22/w... | For a bit of background, from back in May: religionnews.com/2026/05/22/w... | ↗Not briefed |
bluesky bluesky-ai | 1mo | Genuinely engrossing reporting here from the Times. Some of this is a re-tread of stuff that has already been reported, but the extensive … | Genuinely engrossing reporting here from the Times. Some of this is a re-tread of stuff that has already been reported, but the exten… | ↗Not briefed |
bluesky bluesky-ai | 1mo | It would be an unfathomable job to try to put my life and responsibilities into lists for an LLM to attempt to manage and even if I did tha… | It would be an unfathomable job to try to put my life and responsibilities into lists for an LLM to attempt to manage and even if I di… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Every discerning human should take potential dangers & benefits of AI seriously, do not dismiss it. The technology is life changing, positi… | Every discerning human should take potential dangers & benefits of AI seriously, do not dismiss it. The technology is life changing, p… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Derivative games from your hard work: Artificial intelligence derivation code tools: Turning your hard work into someone else's profits. … | Derivative games from your hard work: Artificial intelligence derivation code tools: Turning your hard work into someone else's profi… | ↗Not briefed |
hn hn-ai | 1mo | How the OpenAI DNS Jailbreak Worked | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | ok who's the a.i. vendor providing the llm that powers this chatbot | ok who's the a.i. vendor providing the llm that powers this chatbot | ↗Not briefed |
arxiv arxiv-ai | 1mo | Predictive Geometry of Hidden Trajectories in Transformers | Decoder-only transformers are trained only through a terminal next-token prediction loss, yet this loss constrains every intermediate… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Volatility-Clustering Adaptation for Financial Time Series | Time-series foundation models are increasingly adapted to new domains through fine-tuning on target data, under the implicit assumptio… | ↗Not briefed |
rss the-decoder | 1mo | ElevenLabs' new v4 speech model makes AI voices more expressive and consistent🔧 ElevenLabs | Elevenlabs' new speech model, Eleven v4, follows cues for laughter and whispering more accurately and keeps voices consistent across l… | ↗Not briefed |
rss techmeme | 1mo | Miami-based Jeeves, which launched stablecoin-based payments earlier this year, raised a $110M Series C led by CoinFund, with participation from a16z and others (Maria Clara Cobo/Bloomberg) | Maria Clara Cobo / Bloomberg: Miami-based Jeeves, which launched stablecoin-based payments earlier this year, raised a $110M Series C… | ↗Not briefed |
bluesky bluesky-ai | 1mo | My first read is that like many other NDS websites the real functionality is ‘coming soon’ - I tested a VA disability claim and it directs … | My first read is that like many other NDS websites the real functionality is ‘coming soon’ - I tested a VA disability claim and it dir… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Billiger.de Products: A Bilingual Entity Matching Benchmark🔧 GPT-5 | Existing product matching benchmarks primarily contain English-language product data and are often dominated by a single product categ… | ↗Not briefed |
arxiv arxiv-ai | 1mo | PolyOCR-Venus: Unified OCR Foundation Models for Text-Centric Visual Intelligence | Optical Character Recognition (OCR) is evolving from plain-text transcription toward general visual intelligence, requiring models to… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Anthropic IPO filing warns AI may pose 'existential risks to humanity' ➡️ https://l.euronews.com/ZYsy | Anthropic IPO filing warns AI may pose 'existential risks to humanity' ➡️ https://l.euronews.com/ZYsy | ↗Not briefed |
bluesky bluesky-ai | 1mo | Anthropic says its AI models pose ‘existential risk to humanity’ in leaked IPO filing: report https://fascinating.news/r/5ts6jcGr | Anthropic says its AI models pose ‘existential risk to humanity’ in leaked IPO filing: report https://fascinating.news/r/5ts6jcGr | ↗Not briefed |
arxiv arxiv-ai | 1mo | VIF-Bench: Evaluating Visual Instruction Following in Multi-Reference Image Generation | Recent multimodal image generation models can take multiple images and textual instructions as input, enabling reference-based generat… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Generative Interactions: Weaving Multiparty Human Motion with Bilevel Latent Dynamics | Human social behaviour is not a collection of independent motions, but a jointly organised process in which group dynamics and individ… | ↗Not briefed |
hn hn-ai | 1mo | Don't Attribute Intelligence to It | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | ChatGPT maker OpenAI cancels release of newest AI model due to safety concerns It is madness investing trillions in a tech that is fueling …🔧 ChatGPT | ChatGPT maker OpenAI cancels release of newest AI model due to safety concerns It is madness investing trillions in a tech that is fue… | ↗Not briefed |
bluesky bluesky-ai | 1mo | agente de ia que invade a sua casa e guarda toda a pilha de roupa a anthropic não inventa né. engraçado. | agente de ia que invade a sua casa e guarda toda a pilha de roupa a anthropic não inventa né. engraçado. | ↗Not briefed |
arxiv arxiv-ai | 1mo | Width Expansion as a Method for Class Incremental Learning | Class Incremental Learning (Class-IL) requires models to learn new classes over time while preserving previously acquired knowledge wi… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Locating Answer-Correctness Signals in Frozen Large Language Models | Language models expose internal signals that predict whether an answer is correct, readable from a single forward pass of a frozen mod… | ↗Not briefed |
hn hn-ai | 1mo | Windows and macOS both use thread slot 0x20, so fiber games crash under Wine | — | ↗Not briefed |
arxiv arxiv-ai | 1mo | GARDiff: Graph-Aligned Residual Diffusion for Probabilistic Multivariate Time-Series Forecasting | Diffusion models have recently shown strong potential for probabilistic multivariate time-series forecasting by modeling complex condi… | ↗Not briefed |
hn hn-ai | 1mo | Claude Code saves transcript.jsonl. The log is the truth | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | Lawyers are one of the few professionals you genuinely could replace with an LLM and get better results, they're normally just bullshit mer… | Lawyers are one of the few professionals you genuinely could replace with an LLM and get better results, they're normally just bullshi… | ↗Not briefed |
hn hn-ai | 1mo | I let AI agents write most of a live-trading codebase | — | ↗Not briefed |
arxiv arxiv-ai | 1mo | Reader Proficiency Shapes Layer-wise Surprisal Profiles | Reading behaviour varies not only with linguistic input, but also with reader proficiency. In this study, we investigate whether the l… | ↗Not briefed |
arxiv arxiv-ai | 1mo | WISE-ATTA: When to Ask for Labels in Budgeted Active Test-Time Adaptation | Active test-time adaptation (ATTA) improves robustness under distribution shift by updating a deployed model during inference while se… | ↗Not briefed |
arxiv arxiv-ai | 1mo | EngiWorld: What Can Frontier Agents Deliver in Professional Engineering Environments? | Autonomous agents have made rapid progress in general-purpose computer use, but reliable automation of professional industrial enginee… | ↗Not briefed |
hn hn-ai | 1mo | macOS Golden Gate Is a Buggy Mess | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | JUST IN: Anthropic's Claude AI is currently down for many users worldwide. | JUST IN: Anthropic's Claude AI is currently down for many users worldwide. | ↗Not briefed |
arxiv arxiv-ai | 1mo | Learning from Shared-Control Overrides: Context-Driven Acceleration Profile Prediction for Personalized Overtaking | Adaptive Cruise Control (ACC) systems are typically calibrated for an average driver, often resulting in a mismatch between vehicle be… | ↗Not briefed |
hn hn-ai | 1mo | Microsoft Azure OpenAI Service Down in EU | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | ”AI is going to kill us all you so you should give us a lot of money“ isn‘t like a new ploy for these guys, it’s in their bones. It was the… | ”AI is going to kill us all you so you should give us a lot of money“ isn‘t like a new ploy for these guys, it’s in their bones. It wa… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I don’t think a lot of people are aware that OpenAI and Anthropic are making AI while dooming about AI because: - OpenAI started as an “AI … | I don’t think a lot of people are aware that OpenAI and Anthropic are making AI while dooming about AI because: - OpenAI started as an… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I've been seeing quite a few posts on socials about how powerful Claude Opus 5.5 by @anthropic.com is, in terms of replicating game code, b… | I've been seeing quite a few posts on socials about how powerful Claude Opus 5.5 by @anthropic.com is, in terms of replicating game co… | ↗Not briefed |
arxiv arxiv-ai | 1mo | When Models Don't Manipulate Manifolds: The Geometry of a Comparison Task | One of the current premises of mechanistic interpretability research is that detailed accounts of the geometry of neural network repre… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Learning Expressive and Compositional Motion Representation via Spectral Skills | Robotic foundation models offer a promising path toward general-purpose humanoid robot control, often through hierarchical architectur… | ↗Not briefed |