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 |
|---|---|---|---|---|
rss techcrunch-ai | 2d | Nvidia says its open-source AI strategy remains intact after Jensen Huang’s exit🔧 Hugging Face | Nvidia's interim CEO reaffirmed its open-source AI strategy and plans to acquire Hugging Face, despite Jensen Huang's sudden departure. | ↗Not briefed |
arxiv arxiv-ai | 2d | RefinementGuard: Counterexample-Directed Semantic Qualification for Hidden Agent-Tool Execution | Tool runtimes can lower an agent-selected operation into a hidden parallel graph, reducing latency and context while risking semantic… | ↗Not briefed |
bluesky bluesky-ai | 2d | So uh... is it even possible to get a book out into the world now and keep it away from some AI model or another? As I look at a couple of… | So uh... is it even possible to get a book out into the world now and keep it away from some AI model or another? As I look at a coup… | ↗Not briefed |
bluesky bluesky-ai | 2d | It really isn’t stated enough that if you’re using ai you’re poisoning human beings. If someone thinks their open model is free of it they’… | It really isn’t stated enough that if you’re using ai you’re poisoning human beings. If someone thinks their open model is free of it… | ↗Not briefed |
arxiv arxiv-ai | 2d | Optical Neural Networks Should Be Evaluated by Function, Not Only Physics | Optical neural networks (ONNs) are attracting significant attention as alternative platforms for machine learning (ML). In an ONN, an… | ↗Not briefed |
bluesky bluesky-ai | 2d | I predict Anthropic is gonna drop the price some on Sonnet. Maybe 10%-20%. I feel like it's overpriced. I mean conspicuously. Oh. Hey. Sor… | I predict Anthropic is gonna drop the price some on Sonnet. Maybe 10%-20%. I feel like it's overpriced. I mean conspicuously. Oh. Hey… | ↗Not briefed |
arxiv arxiv-ai | 2d | Label-Free Role-Aware Segmentation of Classroom Recordings: Recovering the Teacher from Transferable Structure | How much the teacher talks and when are among the most informative and costly-to-measure signals of teaching. Yet, classroom audio is… | ↗Not briefed |
arxiv arxiv-ai | 2d | JEPA as a Transferable Optimizer: Learning and Understanding Black-Box Optimization from Reward-Agnostic Pretraining | Black-box optimization evaluates candidates through simulations, rollouts, or experiments, often without gradients. Learned optimizers… | ↗Not briefed |
bluesky bluesky-ai | 2d | んあ…おはよう…♪ 今日もがんばりましょう♡ 【 参加者募集中 】 ブログ・日記・イラスト・写真・小説をBlueskyアカウントで投稿・販売できる複合型コンテンツサービス ➡️ https://atelier.fan #ブログ #日記 #イラスト #写真 #小説 #AI… | んあ…おはよう…♪ 今日もがんばりましょう♡ 【 参加者募集中 】 ブログ・日記・イラスト・写真・小説をBlueskyアカウントで投稿・販売できる複合型コンテンツサービス ➡️ https://atelier.fan #ブログ #日記 #イラスト #写真 #小説… | ↗Not briefed |
bluesky bluesky-ai | 2d | Funny... OpenAI announces security products, then all hell breaks loose. Hmmmmmm..... I smell procurement fraud, Fed griftomatic style. … | Funny... OpenAI announces security products, then all hell breaks loose. Hmmmmmm..... I smell procurement fraud, Fed griftomatic st… | ↗Not briefed |
hn hn-ai | 2d | I hate GPT-6.2 Astra | — | ↗Not briefed |
bluesky bluesky-ai | 2d | Lmfao the kids just told me that Google search doesn't work for them anymore and they have to use some knock off search (because Google dis… | Lmfao the kids just told me that Google search doesn't work for them anymore and they have to use some knock off search (because Googl… | ↗Not briefed |
arxiv arxiv-ai | 2d | FutureBench: A Renewable Framework for Evaluating Knowledge-Context Conflict in LLMs | Asking a model to check retrieved text can also tell it which source to trust. We introduce FutureBench, a renewable framework for eva… | ↗Not briefed |
bluesky bluesky-ai | 2d | Green satin, red stockings, zero rush. The uncensored Naughty Christmas set lives on Fansly. Real model, no AI, 18+. Black November: 35% of… | Green satin, red stockings, zero rush. The uncensored Naughty Christmas set lives on Fansly. Real model, no AI, 18+. Black November: 3… | ↗Not briefed |
bluesky bluesky-ai | 2d | Wealthsimple, I like you dawg, but stop asking me if I want in on the Anthropic IPO, does my TFSA full of VEQT scream IPO goblin to you? | Wealthsimple, I like you dawg, but stop asking me if I want in on the Anthropic IPO, does my TFSA full of VEQT scream IPO goblin to yo… | ↗Not briefed |
arxiv arxiv-ai | 2d | Entropy-Regularized Integrated Gradients for Stable Feature Attribution | Path-based attribution methods like Integrated Gradients (IG) are widely used to interpret neural networks, yet imperceptible input pe… | ↗Not briefed |
arxiv arxiv-ai | 2d | DuoCL: Dual-Path Online Continual Learning | Online continual learning (OCL) requires learning from a data stream whose arrival rate is independent of model training speed. When t… | ↗Not briefed |
rss simon-willison | 2d | External sandboxes for coding agents | — | ↗Not briefed |
hn hn-ai | 2d | Ask HN: I read Blake Masters' notes for Zero to One went viral on HN | What exactly about the notes made them go viral on HN? Also, I would like to thank this community for getting that book published; it… | ↗Not briefed |
rss marktechpost | 2d | Perplexity Computer Adds Multiagent: Up to 16 AI Agents That Cross-Check Work | Perplexity added Multiagent to Computer, an effort level above Ultra where a lead agent runs up to 16 agents on Claude, GPT and Gemini… | ↗Not briefed |
arxiv arxiv-ai | 2d | Efficient Variance-Aware Reinforcement Learning in Time-Inhomogeneous Linear MDPs | We study variance-dependent reinforcement learning in episodic time-inhomogeneous linear Markov decision processes (MDPs) with feature… | ↗Not briefed |
rss techmeme | 2d | Sources detail how the Microsoft 365 outage began in Switzerland and spread as failover systems shifted traffic, leaving 80%+ of Exchange servers unresponsive (Tom Warren/The Verge) | Tom Warren / The Verge: Sources detail how the Microsoft 365 outage began in Switzerland and spread as failover systems shifted traffi… | ↗Not briefed |
arxiv arxiv-ai | 2d | GRAFT: Graph-Regularized Prototype Fields for Realistic Test-Time Adaptation of VLMs | Unlabeled test data can improve the zero-shot classifier of a vision-language model, yet transductive and test-time adaptation methods… | ↗Not briefed |
bluesky bluesky-ai | 2d | OpenAI is releasing a new version of its most advanced Astra AI model nearly two months after the company scrapped a prior update to establ… | OpenAI is releasing a new version of its most advanced Astra AI model nearly two months after the company scrapped a prior update to e… | ↗Not briefed |
bluesky bluesky-ai | 2d | “Surely we need to update the law so we can prosecute the officers of tech companies like OpenAI for serious misconduct?” writes Toby Walsh… | “Surely we need to update the law so we can prosecute the officers of tech companies like OpenAI for serious misconduct?” writes Toby… | ↗Not briefed |
bluesky bluesky-ai | 2d | the purpose of a system is what it does (getting nuked by an llm) | the purpose of a system is what it does (getting nuked by an llm) | ↗Not briefed |
rss techcrunch-ai | 2d | Nvidia says it doesn’t plan to reappoint Jensen Huang as CEO | The former CEO offered to return to his top role at the company after his resignation last week. | ↗Not briefed |
bluesky bluesky-ai | 2d | I applied to see how hard it would be, and Anthropic gave me access to Mythos. I would’ve felt better if they hadn’t 🤣 Haven’t tried it,… | I applied to see how hard it would be, and Anthropic gave me access to Mythos. I would’ve felt better if they hadn’t 🤣 Haven’t trie… | ↗Not briefed |
arxiv arxiv-ai | 2d | ARISE: Practical Self-Evolving Memory for Accurate and Cost-Efficient Enterprise Agents | Enterprise LLM agents operate in dynamic environments where foundation models evolve rapidly, workflows update frequently, and deploym… | ↗Not briefed |
rss simon-willison | 2d | GPT‑6.2 Astra and GPT‑6.2 Sol | — | ↗Not briefed |
hn hn-ai | 2d | GTA 6 on Google: four years of searches, in numbers | — | ↗Not briefed |
bluesky bluesky-ai | 2d | ai’s economic value: all-time high trust in ai: all-time low not sustainable | ai’s economic value: all-time high trust in ai: all-time low not sustainable | ↗Not briefed |
hn hn-ai | 2d | Show HN: Infoboxing – Guess the Wikipedia article from its infobox | I made a daily game where you try to identify five Wikipedia articles by looking at their infoboxes with the name and picture hidden.… | ↗Not briefed |
hn hn-ai | 2d | Show HN: Parallel Search Eval Harness | — | ↗Not briefed |
rss arstechnica-tech | 2d | Hundreds of websites deliver potent malware cocktail to unsuspecting visitors | Malware cocktail delivers cryptocurrency theft payloads and iOS exploits, among other things. | ↗Not briefed |
bluesky bluesky-ai | 2d | Amazon and Google will together own nearly 30% of Anthropic after its IPO | Amazon and Google will together own nearly 30% of Anthropic after its IPO | ↗Not briefed |
hn hn-ai | 2d | Show HN: DecksMD – local first, Markdown flashcards app and plugin for Obsidian | I built a flashcards app and plugin for obsidian, based on markdown syntax. You use headers and paragraphs, tables or highlighted text… | ↗Not briefed |
hn hn-ai | 2d | Show HN: Sikemux – a whole team of AI coders in one window (Rust, MIT) | Here's why I started building Sikemux:<p>At first I was using Cursor. The problem with it was, I had to open a new window for eac… | ↗Not briefed |
rss techcrunch-ai | 2d | Amazon and Google will together own nearly 30% of Anthropic after its IPO | Amazon will own just under 15% of Anthropic after the IPO, with the option to go much higher. Google is a close second with holdings o… | ↗Not briefed |
arxiv arxiv-ai | 2d | ExiST: Existence-Aware Self-Training for Semi-Supervised Multi-Organ Segmentation🔧 Consensus | Semi-supervised multi-organ segmentation should learn more than a dozen organs from a handful of labeled scans, and the smallest organ… | ↗Not briefed |
hn hn-ai | 2d | Show HN: dead-tests finds tests that never run or can never fail | — | ↗Not briefed |
bluesky bluesky-ai | 2d | I mean, also, it’s the week before Black Friday. It makes sense that big organizations would want to upgrade or stress test systems now. … | I mean, also, it’s the week before Black Friday. It makes sense that big organizations would want to upgrade or stress test systems no… | ↗Not briefed |
bluesky bluesky-ai | 2d | What I would actually like would be to force everyone to write inpatient notes in standard SOAP and clinic notes in equivalent formats, and… | What I would actually like would be to force everyone to write inpatient notes in standard SOAP and clinic notes in equivalent formats… | ↗Not briefed |
hn hn-ai | 2d | Browser-only choropleth map editor | — | ↗Not briefed |
bluesky bluesky-ai | 2d | The Four Horsemen of the AI-pocalypse: Claude, Gemini, ChatGPT, and Grok.🔧 ChatGPT | The Four Horsemen of the AI-pocalypse: Claude, Gemini, ChatGPT, and Grok. | ↗Not briefed |
rss wired-ai | 2d | Nvidia’s Board Has No Plans to Bring Back Jensen Huang | Despite pressure from President Trump, Elon Musk, and Silicon Valley executives, Nvidia’s board says it “does not intend” to reinstate… | ↗Not briefed |
hn hn-ai | 2d | LiberNT: A Free Windows-Compatible Operating System Targeting Windows11 | — | ↗Not briefed |
hn hn-ai | 2d | Latent Lattice Language Models | — | ↗Not briefed |
arxiv arxiv-ai | 2d | Receiver-Conditioned Clarification for Multi-Agent Food Assistance | Food insecurity creates a pressing need for guidance that helps households find usable food resources and assess food choices in light… | ↗Not briefed |
arxiv arxiv-ai | 2d | Safe to Omit, Worth Omitting? Keeping the Answer and Cutting Latency When an LLM Agent Skips Another's Message | In multi-agent systems, a receiver agent often waits for a helper agent's message (text or its latent KV cache), which can improve per… | ↗Not briefed |