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 | 7d | having a very "oh, you're concerned about poor-quality LLM output? walk me through your practices for evaluating query plans" morning (it … | having a very "oh, you're concerned about poor-quality LLM output? walk me through your practices for evaluating query plans" morning… | ↗Not briefed |
bluesky bluesky-ai | 7d | Zwei OpenAI-Agenten haben aktiv Sicherheitswarnungen unterdrückt. Es wird geprüft, ob dies aus Missverständnis oder Absicht geschah. #OpenAI | Zwei OpenAI-Agenten haben aktiv Sicherheitswarnungen unterdrückt. Es wird geprüft, ob dies aus Missverständnis oder Absicht geschah. #… | ↗Not briefed |
arxiv arxiv-ai | 7d | Same Luck, Different Choice: Twin-World Credit Assignment for Agentic Reinforcement Learning | Agentic reinforcement learning rewards a complete episode, yet the policy needs credit for every decision inside it. Counterfactual br… | ↗Not briefed |
bluesky bluesky-ai | 7d | AI lobbyists had a busy summer www.theverge.com/policy/10165... | AI lobbyists had a busy summer www.theverge.com/policy/10165... | ↗Not briefed |
hn hn-ai | 7d | Show HN: JobTrace – a self-hosted job tracker with Telegram alerts | — | ↗Not briefed |
bluesky bluesky-ai | 7d | If we're going to impose marketshare liability, why impose it on Anthropic and OpenAI and not GPU manufacturers? | If we're going to impose marketshare liability, why impose it on Anthropic and OpenAI and not GPU manufacturers? | ↗Not briefed |
hn hn-ai | 7d | One setjmp call was eating 81% of Geolith's emulator frame time | I'm the developer of NeoBox (https://neo-box.app). It's a Neo Geo emulator frontend for Mac + iPhone + iPad as one… | ↗Not briefed |
rss techcrunch-ai | 7d | Chinese hackers spied on thousands of companies after breaching IT giant ConnectWise | Microsoft points the blame for ConnectWise's breach at a China-aligned hacker, which relied on AI agents to rapidly hack into thousand… | ↗Not briefed |
bluesky bluesky-ai | 7d | But asbestos was a legitimate product we had reliable marketshare information about. I don't think we have any reason to believe that ha… | But asbestos was a legitimate product we had reliable marketshare information about. I don't think we have any reason to believe th… | ↗Not briefed |
arxiv arxiv-ai | 7d | PYTHIA, a transferable machine learning framework for simultaneous defect and adsorption site optimization | Machine learning has permeated material optimization, providing high value. Nevertheless, its re- liance on a high number of samples l… | ↗Not briefed |
arxiv arxiv-ai | 7d | LACToS: Local Anomaly Detection via Conditional Typicality of Superpixels | Local anomaly detection is the task of identifying spatially confined anomalous or out-of-distribution (OOD) regions within an otherwi… | ↗Not briefed |
bluesky bluesky-ai | 7d | This website has been stupid enough about LLMs to radicalise me into being pro-LLM frankly | This website has been stupid enough about LLMs to radicalise me into being pro-LLM frankly | ↗Not briefed |
bluesky bluesky-ai | 7d | Everwind also does not pay their taxes- according to the province of Newfoundland where their ‘green energy’ project failed to make the mon… | Everwind also does not pay their taxes- according to the province of Newfoundland where their ‘green energy’ project failed to make th… | ↗Not briefed |
bluesky bluesky-ai | 7d | Millions of agents are just assailing the internet rn bc of the php issue. probably impossible to identify which companies originate the ag… | Millions of agents are just assailing the internet rn bc of the php issue. probably impossible to identify which companies originate t… | ↗Not briefed |
bluesky bluesky-ai | 7d | Filings: Nvidia, Anthropic, and OpenAI had their highest-spending lobbying quarter between July and September, spending $2.2M, $2.1M, and $… | Filings: Nvidia, Anthropic, and OpenAI had their highest-spending lobbying quarter between July and September, spending $2.2M, $2.1M,… | ↗Not briefed |
bluesky bluesky-ai | 7d | "The day one reaction was like, 'We're cooked, and our field is over,'" OpenAI's CEO said. "'We're telling our grad students to switch majo… | "The day one reaction was like, 'We're cooked, and our field is over,'" OpenAI's CEO said. "'We're telling our grad students to switch… | ↗Not briefed |
rss techmeme | 7d | Filings: Nvidia, Anthropic, and OpenAI had their highest-spending lobbying quarter between July and September, spending $2.2M, $2.1M, and $1.5M, respectively (Lauren Feiner/The Verge) | Lauren Feiner / The Verge: Filings: Nvidia, Anthropic, and OpenAI had their highest-spending lobbying quarter between July and Septemb… | ↗Not briefed |
bluesky bluesky-ai | 7d | DHS ordered federal civilian agencies to stop buying and to remove DeepSeek technology. #DeepSeek #Cybersecurity #DHS #CISA #TechPolicy #A… | DHS ordered federal civilian agencies to stop buying and to remove DeepSeek technology. #DeepSeek #Cybersecurity #DHS #CISA #TechPoli… | ↗Not briefed |
arxiv arxiv-ai | 7d | An Empirical Study of Recent Large Language Models for Knowledge Graph to Text Generation on Standard Benchmarks | Knowledge graph-to-text (KG-to-text) generation verbalizes a set of knowledge graph triples into fluent natural language. While fine-t… | ↗Not briefed |
arxiv arxiv-ai | 7d | Direct Inference on Admissible Sets (Diona) | We propose a new regression method named Direct Inference on Admissible Sets (Diona) that directly uses samples to estimate the state… | ↗Not briefed |
arxiv arxiv-ai | 7d | Nonparametric autoregressive estimation of probability distributions with parallel gradient boosting | Nonparametric estimation of joint probability distributions is a core problem in statistics, which is notoriously challenging with hig… | ↗Not briefed |
rss techmeme | 7d | Atlanta-based Tractian, which uses AI and sensors to predict machine failures and recommend maintenance, raised a $130M Series D at a $1.4B valuation (Rocket Drew/The Information) | Rocket Drew / The Information: Atlanta-based Tractian, which uses AI and sensors to predict machine failures and recommend maintenance… | ↗Not briefed |
bluesky bluesky-ai | 7d | DeepSeek raised around $15 billion in one of the China's biggest-ever financing rounds as part of efforts to keep up with Western frontier … | DeepSeek raised around $15 billion in one of the China's biggest-ever financing rounds as part of efforts to keep up with Western fron… | ↗Not briefed |
arxiv arxiv-ai | 7d | Beyond Causal Masks: Auditing and Localizing Hidden Future-Token Leakage in Adaptive Sequence Models | Hidden future-token leakage can create false architectural gains: an autoregressive model may appear substantially better under teache… | ↗Not briefed |
hn hn-ai | 7d | Show HN: Ternlux – Free, local-first, bring-your-own-key AI pipeline builder🔧 Ollama🔧 OpenRouter🔧 Pinecone | For the past 12 years as a data engineer, I've been building pipelines across many different business categories and sizes. Most… | ↗Not briefed |
hn hn-ai | 7d | Anthropic builds its own source control system | — | ↗Not briefed |
arxiv arxiv-ai | 7d | Jev-DLC: Decision-Level Continual Learning from Typed Feedback | Individual preferences often become visible only after deployment, when a user accepts one decision or revokes it in favor of another.… | ↗Not briefed |
arxiv arxiv-ai | 7d | The Geometry of Metaphor: Probing and Steering Metaphor Representations in Large Language Models | Metaphor is a crucial aspect of human cognition. As such, we would like to understand how it is represented in large language models (… | ↗Not briefed |
bluesky bluesky-ai | 7d | The infrastructure of the internet is full of mistakes; basically everything is slapped together with duct tape and hope. OpenAI has decide… | The infrastructure of the internet is full of mistakes; basically everything is slapped together with duct tape and hope. OpenAI has d… | ↗Not briefed |
bluesky bluesky-ai | 7d | The jokes you can make about "the only way to stop a bad guy with an llm is a good guy with an llm" are endless though | The jokes you can make about "the only way to stop a bad guy with an llm is a good guy with an llm" are endless though | ↗Not briefed |
arxiv arxiv-ai | 7d | From Prompt Rules to Golden Data Model: Production-Scale Generation of Personalized, Compliant Pharmaceutical Claims | Large language models (LLMs) are increasingly used to generate domain-specific commercial & medical content (e.g., claims, statements,… | ↗Not briefed |
rss marktechpost | 7d | Cohere Releases Rerank 5 With Passage-Level Context Focus and Calibrated Scores🔧 Cohere | Cohere Rerank 5 ships in Pro and Fast tiers with a 32K context window, Context Focus passage extraction and calibrated relevance score… | ↗Not briefed |
bluesky bluesky-ai | 7d | Goldman Sachs launches its Small Business AI Institute as part of a $500 million expansion of its entrepreneur support program, partnering … | Goldman Sachs launches its Small Business AI Institute as part of a $500 million expansion of its entrepreneur support program, partne… | ↗Not briefed |
hn hn-ai | 7d | Post-training 4B model with RL to fly Tesla Roadster 21% better than GPT-6 Astra | — | ↗Not briefed |
arxiv arxiv-ai | 7d | Towards User Simulators That Stay On Task And Still Sound Natural | Users increasingly rely on AI systems for extended, multi-turn tasks, yet evidence indicates that LLM-based system performance degrade… | ↗Not briefed |
hn hn-ai | 7d | AgentIntersect World – local-first 3D workspace for your coding agents | — | ↗Not briefed |
arxiv arxiv-ai | 7d | DuoCook: An Overcooked-Inspired Benchmark for Visual Planning and Coordination | Coordinating two players in a cooking task requires a vision-language agent to track changing scene states, sequence recipe steps, and… | ↗Not briefed |
bluesky bluesky-ai | 7d | I don’t want to give them credit for finding it because I do not believe they found anything. It was an AI model that “found” it. | I don’t want to give them credit for finding it because I do not believe they found anything. It was an AI model that “found” it. | ↗Not briefed |
arxiv arxiv-ai | 7d | Piano2Pop: Expanding Piano Demos into Pop Productions Without Aligned Supervision | We introduce Piano2Pop, a music-expansion task that transforms piano MIDI into instrumental pop productions. Enabled by recent advance… | ↗Not briefed |
rss techmeme | 7d | METR says a pilot test of OpenAI's monitoring system found vulnerabilities, including an exploit that could have let AI agents hide hundreds of recent actions (Anzar Mehraj/Reuters) | Anzar Mehraj / Reuters: METR says a pilot test of OpenAI's monitoring system found vulnerabilities, including an exploit that could ha… | ↗Not briefed |
hn hn-ai | 7d | F3D 4.0.0 adds FreeCAD (.FCStd) support | — | ↗Not briefed |
arxiv arxiv-ai | 7d | FairFRESH: A Neuro-Symbolic Framework for Fair and Explainable Household Food Recommendation | Recommender systems typically serve one user at a time. However, in households, members share a finite pantry and follow different die… | ↗Not briefed |
hn hn-ai | 7d | Open AI vs. frontier models: the hidden cost of your data (2026) | — | ↗Not briefed |
hn hn-ai | 7d | What Doesn't Get Cheaper | — | ↗Not briefed |
hn hn-ai | 7d | Introducing the Gemini API CLI | — | ↗Not briefed |
rss techmeme | 7d | Sources: Elon Musk has invested roughly $1B in Skydance; in 2022, David Ellison's father, Larry Ellison, invested $1B in Musk's deal to take X private (Rohan Goswami/Semafor) | Rohan Goswami / Semafor: Sources: Elon Musk has invested roughly $1B in Skydance; in 2022, David Ellison's father, Larry Ellison, inve… | ↗Not briefed |
hn hn-ai | 7d | Show HN: ForgeOps, error tracking with a mode where user data stays in your app | — | ↗Not briefed |
arxiv arxiv-ai | 7d | Structure-preserving reduced-order models of thin panels in supersonic and hypersonic flow | We build structure-preserving reduced-order models for both challenge problems of the AePW-4 High-Speed Working Group. They are a clam… | ↗Not briefed |
hn hn-ai | 7d | PCI Scope Lens: see which checkout scripts put you in PCI scope | — | ↗Not briefed |
bluesky bluesky-ai | 7d | Should Nova Scotia be partnering with multinational corporations to support US tech giant Anthropic at the expense of our kid’s health and … | Should Nova Scotia be partnering with multinational corporations to support US tech giant Anthropic at the expense of our kid’s health… | ↗Not briefed |