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 | 29d | Gemini 4 Argon is now available for developers | — | ↗Not briefed |
bluesky bluesky-ai | 29d | There is a whole section on the coup de grâce---it's to be done by the NCOIC. 2/2 www.loc.gov/resource/llm... | There is a whole section on the coup de grâce---it's to be done by the NCOIC. 2/2 www.loc.gov/resource/llm... | ↗Not briefed |
bluesky bluesky-ai | 29d | The fourth-generation Apple TV 4K is here, complete with Siri AI enhancements and Find My on the Siri Remote. Here's how much difference fo… | The fourth-generation Apple TV 4K is here, complete with Siri AI enhancements and Find My on the Siri Remote. Here's how much differen… | ↗Not briefed |
arxiv arxiv-ai | 29d | Model-Based Online Decision Making via Generative Trajectory Planning | Online decision-making faces a tension between efficient inference and long-horizon reasoning. Value-based methods rely on short-horiz… | ↗Not briefed |
hn hn-ai | 29d | Failing to give Meta Muse a soul | — | ↗Not briefed |
bluesky bluesky-ai | 29d | Translation: Anthropic wrote code that allowed or caused this | Translation: Anthropic wrote code that allowed or caused this | ↗Not briefed |
hn hn-ai | 29d | Frugal Computing Revisited | — | ↗Not briefed |
hn hn-ai | 29d | Show HN: Lowlight – Dim your MacBook keyboard backlight below the macOS minimum | I work in a dark room a lot, and even the lowest keyboard backlight setting on my MacBook was too bright. I noticed macOS fades the ba… | ↗Not briefed |
arxiv arxiv-ai | 29d | From Relation-Specific Heads to Shared Interventions: Contextual Entrainment in Language Models | Large language models rely on prompt context to perform in-context learning, follow formats, and adapt to local evidence. This ability… | ↗Not briefed |
hn hn-ai | 29d | Chinese cheap LLMs could trigger trillion-dollar AI infrastructure reset | — | ↗Not briefed |
bluesky bluesky-ai | 29d | Everyone point and laugh at The Hill headline writers | Everyone point and laugh at The Hill headline writers | ↗Not briefed |
bluesky bluesky-ai | 29d | anthropic is lying through their teeth man, ya‘ll don’t actually believe them | anthropic is lying through their teeth man, ya‘ll don’t actually believe them | ↗Not briefed |
bluesky bluesky-ai | 29d | The Eighteenth Brumaire of OpenAI | The Eighteenth Brumaire of OpenAI | ↗Not briefed |
hn hn-ai | 29d | Visual Category Theory Studio | — | ↗Not briefed |
bluesky bluesky-ai | 29d | 7 engineers wrote a custom vcs + review tool and migrated to it in five months you can sure do neat stuff when you buy tokens by the truck… | 7 engineers wrote a custom vcs + review tool and migrated to it in five months you can sure do neat stuff when you buy tokens by the… | ↗Not briefed |
hn hn-ai | 29d | Smash 64 Recomp Developer kicked for using AI (me) | — | ↗Not briefed |
arxiv arxiv-ai | 29d | SMI: Ontology-Grounded Semantic IDs for Hierarchy-Aware Medical Concept Representation | Biomedical language models (LMs) provide strong representations for clinical tasks, but they often fail to capture the hierarchical st… | ↗Not briefed |
bluesky bluesky-ai | 29d | Yes, but arguably, the second you do some post training on it, or allow tool calling, or add CoT, or any of a million other things it stops… | Yes, but arguably, the second you do some post training on it, or allow tool calling, or add CoT, or any of a million other things it… | ↗Not briefed |
bluesky bluesky-ai | 29d | 🎉 Welcome to TerraFrame, a FOSS4G NA Supporter Sponsor! 🎉 They partner with standards bodies and agencies on open-source tools and stand… | 🎉 Welcome to TerraFrame, a FOSS4G NA Supporter Sponsor! 🎉 They partner with standards bodies and agencies on open-source tools and… | ↗Not briefed |
arxiv arxiv-ai | 29d | Known-Voice Cloning and the Limits of Personality Transfer for Character Design | Recent advances in zero-shot voice cloning systems make it possible to reproduce a recognizable voice from a short recording and attac… | ↗Not briefed |
bluesky bluesky-ai | 29d | This sort of behavior is so absurdly anti-intellectual. Yeah, sure, massively downvote me for correctly answering a question being asked. B… | This sort of behavior is so absurdly anti-intellectual. Yeah, sure, massively downvote me for correctly answering a question being ask… | ↗Not briefed |
hn hn-ai | 29d | Show HN: Get your license to do math in 5 minutes | — | ↗Not briefed |
bluesky bluesky-ai | 29d | The ambiguity of AI's economics is a weapon, and our best defense against it is to refuse to accept bullshit metrics and numbers as gospel.… | The ambiguity of AI's economics is a weapon, and our best defense against it is to refuse to accept bullshit metrics and numbers as go… | ↗Not briefed |
bluesky bluesky-ai | 29d | Four years in, we have no understanding of the economics of AI — a terrifying prospect considering the amount of money in play. This ambig… | Four years in, we have no understanding of the economics of AI — a terrifying prospect considering the amount of money in play. This… | ↗Not briefed |
bluesky bluesky-ai | 29d | In a vacuum of good information, bullish projections and forecasts override good sense, because, for the most part, most journalists and an… | In a vacuum of good information, bullish projections and forecasts override good sense, because, for the most part, most journalists a… | ↗Not briefed |
bluesky bluesky-ai | 29d | Annualized run rates exist only to make bad companies look good. AI law startup Harvey has raised more than $1.5bn in the last year, and ha… | Annualized run rates exist only to make bad companies look good. AI law startup Harvey has raised more than $1.5bn in the last year, a… | ↗Not briefed |
bluesky bluesky-ai | 29d | While I imagine the press is trying, one reason I think we lack the true numbers behind most AI startups is that the financial conditions a… | While I imagine the press is trying, one reason I think we lack the true numbers behind most AI startups is that the financial conditi… | ↗Not briefed |
bluesky bluesky-ai | 29d | It's disturbing to see how little we actually know about how well (or badly) things are going for the AI bubble - hyperscalers won't disclo… | It's disturbing to see how little we actually know about how well (or badly) things are going for the AI bubble - hyperscalers won't d… | ↗Not briefed |
bluesky bluesky-ai | 29d | Annualized run rates are less of a financial metric and more of a marketing tool, albeit a very effective one, allowing AI companies to inf… | Annualized run rates are less of a financial metric and more of a marketing tool, albeit a very effective one, allowing AI companies t… | ↗Not briefed |
bluesky bluesky-ai | 29d | Last week, OpenAI's run rate was revised down by $20bn, crushing the AI trade, despite the fact that run rates are made up numbers with lit… | Last week, OpenAI's run rate was revised down by $20bn, crushing the AI trade, despite the fact that run rates are made up numbers wit… | ↗Not briefed |
arxiv arxiv-ai | 29d | BridgeTwist: Twisting Schrödinger Bridges for Training-Free Conditional Sampling | Generative models are often required to serve many downstream conditional tasks. In the training-free setting, a pretrained unconditio… | ↗Not briefed |
rss techmeme | 29d | Apple says HomeView's facial recognition works only if users have a "registered personal device" with Apple Intelligence and Siri AI on the same Wi-Fi network (John Gruber/Daring Fireball) | John Gruber / Daring Fireball: Apple says HomeView's facial recognition works only if users have a “registered personal device” with A… | ↗Not briefed |
bluesky bluesky-ai | 29d | * AI models may soon self-replicate without access to data centers, Anthropic confesses its crimes against humanity🔧 Replicate | * AI models may soon self-replicate without access to data centers, Anthropic confesses its crimes against humanity | ↗Not briefed |
arxiv arxiv-ai | 29d | Bayesian Test-Time Inference of Task-Aligned Similarity from Weak Interactive Feedback | Language model systems repeatedly make test-time decisions by comparing similarity: they retrieve exemplars, rank candidates, and revi… | ↗Not briefed |
arxiv arxiv-ai | 29d | Texture Interpolation as Texton Matching | Texture interpolation asks what should lie between two textures. Standard methods typically blend pixels or features at corresponding… | ↗Not briefed |
arxiv arxiv-ai | 29d | Kwai Summary Attention V2 Technical Report | Summary-token architectures such as Kwai Summary Attention (KSA) compress distant context into learnable Summary token per chunk, and… | ↗Not briefed |
hn hn-ai | 29d | The Geometric Model of Maintenance | — | ↗Not briefed |
bluesky bluesky-ai | 29d | Google launches Gemini 4 Argon 🤖 Google has released Gemini 4 Argon, claiming it outperforms rival models like GPT-6 Astra and Claude acr… | Google launches Gemini 4 Argon 🤖 Google has released Gemini 4 Argon, claiming it outperforms rival models like GPT-6 Astra and Claud… | ↗Not briefed |
arxiv arxiv-ai | 29d | AMemOps: Evaluating Memory Updates in Long Audio Streams | Audio memory must recognise recurring sources across acoustic changes while keeping their attributes up to date. Overall accuracy does… | ↗Not briefed |
hn hn-ai | 29d | Show HN: SpiderMonkey – System 1 for Code Agents | — | ↗Not briefed |
arxiv arxiv-ai | 29d | Does Pretraining Teach Chemical Language Models Chemistry? | Chemical language models (CLMs) are pretrained on large corpora of molecules with the aim of learning representations that transfer to… | ↗Not briefed |
bluesky bluesky-ai | 29d | I was on hold with anthropic again trying to figure out which of the 3 phone calls and 2 emails I got has the correct information and the p… | I was on hold with anthropic again trying to figure out which of the 3 phone calls and 2 emails I got has the correct information and… | ↗Not briefed |
arxiv arxiv-ai | 29d | Tiny Recursive Message Passing | Message passing neural networks have shown remarkable performance in algorithmic and constraint reasoning, due to their ability to enc… | ↗Not briefed |
arxiv arxiv-ai | 29d | ST-GRU PredRNN-V2: A GRU-Based Reformulation of PredRNN-V2 for Video Prediction | Video prediction requires models to capture complex spatiotemporal dependencies and to retain relevant information over long sequences… | ↗Not briefed |
bluesky bluesky-ai | 29d | Computer Scientist here! This entire fear is alarmist BS to try and push for more data centers. Only way this happens is if Anthropic tell… | Computer Scientist here! This entire fear is alarmist BS to try and push for more data centers. Only way this happens is if Anthropic… | ↗Not briefed |
arxiv arxiv-ai | 29d | MultiAlloc: Budget-Aware Precision Allocation for KV Cache Quantization | The Key-Value (KV) cache constitutes a dominant source of memory consumption during large language model (LLM) inference, particularly… | ↗Not briefed |
arxiv arxiv-ai | 29d | To Align or Not To Align: Check Your COMPASS Before You Train🔧 Modal | Combining pretrained models is a common strategy for building AI systems, whether by aligning a vision encoder with a language model o… | ↗Not briefed |
arxiv arxiv-ai | 29d | Grid or Nodes? Probing Object Memory in World Models Under Partial Observability | A robot's world model must remember objects that leave the camera's view. FloWM, a learned world model, keeps them in a spatial grid o… | ↗Not briefed |
bluesky bluesky-ai | 29d | github.com/pfrazee/pdoo... | github.com/pfrazee/pdoo... | ↗Not briefed |
hn hn-ai | 29d | Xara Xtreme is back. After 20 years. With a vengeance | — | ↗Not briefed |