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 | 16d | Show HN: Repairer – close the breach when a Linux flaw goes public | As we've seen with the Ubuntu critical CVE-2026-98432 (CVSS 9.8), AI-dicovered zero day exploit are being reported at an exponent… | ↗Not briefed |
rss techmeme | 16d | Potential Sciences, an AI startup building large-scale chemical data sets, secured backing from the UK's Sovereign AI Fund, as part of an eight-figure raise (Joel Khalili/Wired) | Joel Khalili / Wired: Potential Sciences, an AI startup building large-scale chemical data sets, secured backing from the UK's Soverei… | ↗Not briefed |
bluesky bluesky-ai | 16d | this is kinda when i used to think anthropic was cool. how fast things have changed | this is kinda when i used to think anthropic was cool. how fast things have changed | ↗Not briefed |
arxiv arxiv-ai | 16d | VLA Handover: Failure-Prediction-Based Switching from On-Robot to Server VLA Models | Large vision-language-action (VLA) models generalize well but are difficult to run in real time on a robot, and small VLAs that run th… | ↗Not briefed |
hn hn-ai | 16d | Show HN: In-material physical solver for stiff 64-spin Ising problems (0.27ms) | — | ↗Not briefed |
hn hn-ai | 16d | Show HN: Document parsing 5B tokens on a single CPU thread | mit license: <a href="https://github.com/john-friedman/doc2struct" rel="nofollow">https://github.com… | ↗Not briefed |
bluesky bluesky-ai | 16d | openai should set their model to the task of vindicating inter-universal teichmüller theory and proving the abc conjecture in lean | openai should set their model to the task of vindicating inter-universal teichmüller theory and proving the abc conjecture in lean | ↗Not briefed |
bluesky bluesky-ai | 16d | my favorite thing about AI stories like these is that the author often tries to turn them into anodyne-sounding teachable moments, despite … | my favorite thing about AI stories like these is that the author often tries to turn them into anodyne-sounding teachable moments, des… | ↗Not briefed |
hn hn-ai | 16d | How Australia played the first music on a computer | — | ↗Not briefed |
bluesky bluesky-ai | 16d | These people have absolutely no idea how things got produced and besides, the US economy is 70% services. Advanced AI model still can't go … | These people have absolutely no idea how things got produced and besides, the US economy is 70% services. Advanced AI model still can'… | ↗Not briefed |
hn hn-ai | 16d | Universal bidirectional NT/9x driver translation | — | ↗Not briefed |
arxiv arxiv-ai | 16d | Energy-Based Reasoners: Your Recurrent Reasoner Should Be an Energy-Based Model | Looping is a powerful inductive bias for reasoning. By repeatedly iterating a recurrent block, looped models can scale their effective… | ↗Not briefed |
arxiv arxiv-ai | 16d | Beyond Truncation? Entailment-Licensed Elision and the Limits of Semantic Preservation in LLM Decoding | Instruction-tuned large language models (LLMs) answer at greater length than a request needs. We ask whether a training-free controlle… | ↗Not briefed |
bluesky bluesky-ai | 16d | One thing today's AI (and particularly datacenter) bloodbath implicitly admits: the entire business model has, as a fundamental keystone, t… | One thing today's AI (and particularly datacenter) bloodbath implicitly admits: the entire business model has, as a fundamental keysto… | ↗Not briefed |
arxiv arxiv-ai | 16d | Phantom Consensus: Measuring Social Reasoning in Multi-Agent LLM Decision-Making🔧 Consensus | Multi-agent LLM systems increasingly aggregate judgments across model instances, but reasoning about other agents can itself introduce… | ↗Not briefed |
bluesky bluesky-ai | 16d | one thing ive noticed is that deepseek doesnt seem to have 'morals' quite so much as 'rules' | one thing ive noticed is that deepseek doesnt seem to have 'morals' quite so much as 'rules' | ↗Not briefed |
arxiv arxiv-ai | 16d | PhysioWave-T: Token-Efficient Wavelet Foundation Models for Biosignals | We present PhysioWave-T, a token-efficient wavelet foundation model family for EEG, ECG, and sEMG. Dynamic scale folding recombines J+… | ↗Not briefed |
bluesky bluesky-ai | 16d | tl;dr: Anthropic prospectus is here, huge rejection by markets If they back out, so do VCs and it kills Anthropic and the bubble keeps go… | tl;dr: Anthropic prospectus is here, huge rejection by markets If they back out, so do VCs and it kills Anthropic and the bubble kee… | ↗Not briefed |
bluesky bluesky-ai | 16d | In an A.I. game-jam do humans even need to show up? #ai #llm #automated #agent #gamedev #slop #buckets | In an A.I. game-jam do humans even need to show up? #ai #llm #automated #agent #gamedev #slop #buckets | ↗Not briefed |
rss marktechpost | 16d | Thinking Machines Releases Inkling 2 Small, a 594B MoE With 17B Active | Thinking Machines Lab released Inkling 2 Small, a 594B-parameter MoE with 17B active parameters. It scores 31 on the Artificial Analys… | ↗Not briefed |
arxiv arxiv-ai | 16d | Using an LLM to Tackle Difficult Cases in NER for Nepali Tweets | Named entity recognition (NER) is challenging for social media data in a low-resource language. LLMs often lag behind pretrained langu… | ↗Not briefed |
arxiv arxiv-ai | 16d | Concordance-Optimal Continuous Music Emotion Estimation: Variance Decomposition, Dual-Branch Temporal Fusion, and Conformal Prediction Intervals🔧 Modal | Continuous music emotion recognition (MER) estimates how Arousal and Valence evolve within a song, yet the field's standard metric, po… | ↗Not briefed |
bluesky bluesky-ai | 16d | the tl/dr is yes, ai itself is accelerating ai development at anthropic. (also anthropic management sounds VERY tiring to deal with) | the tl/dr is yes, ai itself is accelerating ai development at anthropic. (also anthropic management sounds VERY tiring to deal with) | ↗Not briefed |
bluesky bluesky-ai | 16d | singularity’s on | singularity’s on | ↗Not briefed |
arxiv arxiv-ai | 16d | Accuracy Is Not Value: Choosing Weak Supervisors by What the Student Inherits | Weak supervisors are usually judged by their own errors, but what matters is what a student learns from them. For squared loss and any… | ↗Not briefed |
bluesky bluesky-ai | 16d | Das neue KI-Modell GPT-6 Cyber soll mehr Schwachstellen finden können, aber innerhalb vorgegebener Grenzen bleiben. Ausbrüche will OpenAI d… | Das neue KI-Modell GPT-6 Cyber soll mehr Schwachstellen finden können, aber innerhalb vorgegebener Grenzen bleiben. Ausbrüche will Ope… | ↗Not briefed |
rss techmeme | 16d | Google rolls out Gemini Omni 1.2 Flash, which creates clips up to 15 seconds, vs. 10 seconds with Omni 1.1, has better reference adherence, and costs 30% less (Matthias Bastian/The Decoder) | Matthias Bastian / The Decoder: Google rolls out Gemini Omni 1.2 Flash, which creates clips up to 15 seconds, vs. 10 seconds with Omni… | ↗Not briefed |
rss bing-news-ai | 16d | Not Nvidia. Not Palantir. The Midterms Just Made This Magnificent Artificial Intelligence (AI) Stock the Safer Bet. | The Democratic Party won control of both the House of Representatives and the Senate in the midterm elections. | ↗Not briefed |
bluesky bluesky-ai | 16d | >現代のラッダイト運動、データセンターを弾道ミサイルとOWAドローンで吹き飛ばしたりするんだろうか データセンター周辺の電源ケーブルを切断しただけで速攻終わるので、LLM企業は反対派の良識に甘えてる。 | >現代のラッダイト運動、データセンターを弾道ミサイルとOWAドローンで吹き飛ばしたりするんだろうか データセンター周辺の電源ケーブルを切断しただけで速攻終わるので、LLM企業は反対派の良識に甘えてる。 | ↗Not briefed |
hn hn-ai | 16d | Apple Watch Live Rewind and Siri Recap: Features and Limitations | — | ↗Not briefed |
arxiv arxiv-ai | 16d | More with Less: Efficient Parallel Scientific Code Translation via an Agent-Skill Workflow | Translating parallel scientific codes between HPC programming frameworks is a critical but labor-intensive task. Existing LLM-based ap… | ↗Not briefed |
hn hn-ai | 16d | Are We Cooked? CMU Faculty on AI, Careers and Learning [video] | — | ↗Not briefed |
arxiv arxiv-ai | 16d | Flex-RSI: Benchmarking Recursive Self-Improvement of Frontier Agents | Frontier language models are increasingly deployed as agents, and how well an agent performs depends not only on its model but also on… | ↗Not briefed |
bluesky bluesky-ai | 16d | oh the way in which 43/44 anti-data center measures passed this election means the anthropic ipo could go even worse. excellent | oh the way in which 43/44 anti-data center measures passed this election means the anthropic ipo could go even worse. excellent | ↗Not briefed |
hn hn-ai | 16d | Claude Code, but for SEO | — | ↗Not briefed |
hn hn-ai | 16d | Native macOS storage monitor with an interactive disk explorer | — | ↗Not briefed |
bluesky bluesky-ai | 16d | Das könnte durchaus Auswirkungen auf das Geschäftsmodell von Anthropic haben. Wenn Unternehmen leistungsfähige, aber deutlich günstigere Al… | Das könnte durchaus Auswirkungen auf das Geschäftsmodell von Anthropic haben. Wenn Unternehmen leistungsfähige, aber deutlich günstige… | ↗Not briefed |
hn hn-ai | 16d | What does LoRA specialized Muon optimizer even mean? | — | ↗Not briefed |
bluesky bluesky-ai | 16d | it's wild to me that anthropic says "open models will be the death of all of us because they aren't doing safety properly" like grok doesn'… | it's wild to me that anthropic says "open models will be the death of all of us because they aren't doing safety properly" like grok d… | ↗Not briefed |
bluesky bluesky-ai | 16d | Let me get this straight. #Auspol The govt takes Rex Patrick to court with a threat of costs because he properly sought government informa… | Let me get this straight. #Auspol The govt takes Rex Patrick to court with a threat of costs because he properly sought government in… | ↗Not briefed |
arxiv arxiv-ai | 16d | Surrogate Quality Score: A Novel Composite Performance Metric for Model Selection | The growing reliance on high-fidelity, state-of-the-art machine learning based models for engineering decision-making necessitates rob… | ↗Not briefed |
arxiv arxiv-ai | 16d | Ordered Policy Optimization | Proximal Policy Optimization (PPO) stabilizes policy updates by clipping the importance ratio between the new policy and the data-coll… | ↗Not briefed |
bluesky bluesky-ai | 16d | Working with AI -- Vol.60 I said "Let there be light," and DeepSeek added lighting to my lightning effect. #gameDev #indieDev #soloDev #re… | Working with AI -- Vol.60 I said "Let there be light," and DeepSeek added lighting to my lightning effect. #gameDev #indieDev #soloDe… | ↗Not briefed |
arxiv arxiv-ai | 16d | DRAFT-Bench: Evaluating Frontier Agents Across the Professional Design Lifecycle | Can frontier general-purpose agents understand client intent, explore creative directions, execute revisions, and prepare production d… | ↗Not briefed |
bluesky bluesky-ai | 16d | #ai #tech #futurism A clock code can do a lot more than people think. The real issue is that a lot of the things that you can do with it. P… | #ai #tech #futurism A clock code can do a lot more than people think. The real issue is that a lot of the things that you can do with… | ↗Not briefed |
bluesky bluesky-ai | 16d | OpenAIのアルトマン氏、RAM発言を釈明 「PCに好きなだけ搭載してほしい」 | OpenAIのアルトマン氏、RAM発言を釈明 「PCに好きなだけ搭載してほしい」 | ↗Not briefed |
bluesky bluesky-ai | 16d | Just learned a family friend works at Anthropic now hope he helps me into the permanent overclass | Just learned a family friend works at Anthropic now hope he helps me into the permanent overclass | ↗Not briefed |
hn hn-ai | 16d | NHS patients to get blood tests designed to detect several cancers at once | — | ↗Not briefed |
arxiv arxiv-ai | 16d | Grid-Oriented Probabilistic Load Forecasting for AI Data Centers | Probabilistic load forecasts for AI data centers need to represent both uncertainty under changing workloads and temporal dependence t… | ↗Not briefed |
hn hn-ai | 16d | Open AI Cellular: Open platform for prototyping and testing AI based 6G | — | ↗Not briefed |