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 | 1mo | In 2026, OpenAI bots bypass security controls to access multiple Government web sites and extract data. Nobody has been arrested or threate… | In 2026, OpenAI bots bypass security controls to access multiple Government web sites and extract data. Nobody has been arrested or th… | ↗Not briefed |
hn hn-ai | 1mo | Tell HN: OVH price increase for dedicated servers | Since October 1 the OVH price increase also applies to running dedicated servers, at next renewal. A Reddit user posted that their two… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Sentry: Learning to Recover from LLM Agent Failures at Test Time | LLM agents often fail mid-task due to invalid tool calls, repeated actions, or poorly grounded reasoning, and learning from these fail… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Differentiable Koopman Operator for Contrastive Learning on Dynamic Graphs | Real-world interaction networks are inherently dynamic: edges form and dissolve as node behavior shifts over time. Most snapshot-based… | ↗Not briefed |
bluesky bluesky-ai | 1mo | A few months ago, we launched our MCP. And yes, we were the first design tool to do it😉 It connects AI agents directly to Penpot, giving … | A few months ago, we launched our MCP. And yes, we were the first design tool to do it😉 It connects AI agents directly to Penpot, gi… | ↗Not briefed |
arxiv arxiv-ai | 1mo | OLMo-Detect: A Multi-Stage, Confounder-Controlled Benchmark for Membership Inference on Large Language Models | Membership inference on large language models (LLMs) aims to determine whether a given text sample was included in an LLM's training d… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Anthropic invited religious leaders for a meeting about their Ai model Claude, essentially trying to convince them it’s got a soul and dese… | Anthropic invited religious leaders for a meeting about their Ai model Claude, essentially trying to convince them it’s got a soul and… | ↗Not briefed |
bluesky bluesky-ai | 1mo | OpenAIが9月8日に解けたと発表した，数学の難問であるナビエ・ストークス方程式の問題。実は数学者が長年注目してきた問題は依然として解けていないようです。改めてこの難問の要点が何かを巡る議論が起こっており，単なる「AI vs 数学者」の対決を超えた展開になりそう。 nk-sc… | OpenAIが9月8日に解けたと発表した，数学の難問であるナビエ・ストークス方程式の問題。実は数学者が長年注目してきた問題は依然として解けていないようです。改めてこの難問の要点が何かを巡る議論が起こっており，単なる「AI vs 数学者」の対決を超えた展開になりそう。… | ↗Not briefed |
hn hn-ai | 1mo | Woman loses 20 years of 'digital memories' after Meta deletes her accounts | — | ↗Not briefed |
arxiv arxiv-ai | 1mo | PLCWorld: Benchmarking LLM-Generated PLC Programs in Closed-Loop Plant Simulation🔧 GPT-5 | Programmable logic controllers (PLCs) coordinate industrial equipment by reading sensor inputs and issuing control commands. Evaluatin… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Oh joy - another round of filling in that fucking Anthropic Settlement form I've filled in twice already. Getting a sense we're never seein… | Oh joy - another round of filling in that fucking Anthropic Settlement form I've filled in twice already. Getting a sense we're never… | ↗Not briefed |
bluesky bluesky-ai | 1mo | THERE.IS.NO “ROGUE” in AI “OpenAI illegally accesses another NSW govt website” is the heading OpenAI I.P created , curated and deployed… | THERE.IS.NO “ROGUE” in AI “OpenAI illegally accesses another NSW govt website” is the heading OpenAI I.P created , curated and dep… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Unterschrift von Greg Brockman (OpenAI) unter Trumps morally binding KI-Vereinbarung („Unites" States of America). | Unterschrift von Greg Brockman (OpenAI) unter Trumps morally binding KI-Vereinbarung („Unites" States of America). | ↗Not briefed |
rss bing-news-ai | 1mo | Trump to name intelligence chief Jay Clayton as AI czar | President Donald Trump is expected to appoint Jay Clayton, the current US Director of National Intelligence, as his top adviser on art… | ↗Not briefed |
rss bing-news-ai | 1mo | Trump to name intelligence chief Jay Clayton as AI czar | President Donald Trump is expected to appoint Jay Clayton, the current US Director of National Intelligence, as his top adviser on art… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Safeguarding Mutual Correction in Source-Free Domain Adaptation via Cut Statistics | Source-Free Domain Adaptation (SFDA) aims to adapt a source-pretrained model to an unlabeled target domain without access to the origi… | ↗Not briefed |
arxiv arxiv-ai | 1mo | RASPER: Reward-Aligned Summarization of Clinical Notes for EHR Outcome Prediction | Unstructured discharge notes in Electronic Health Records (EHRs) often carry signal complementary to structured medical codes, holding… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Gift Link: “It appeared to Rabbi Navon that Mr. Olah and his team believed that Claude had what philosophers call “moral status” on par wit… | Gift Link: “It appeared to Rabbi Navon that Mr. Olah and his team believed that Claude had what philosophers call “moral status” on pa… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Relevant Evidence Decoding for Audio-Visual Hallucination Mitigation🔧 Modal | Audio-Visual Large Language Models (AV-LLMs) remain prone to cross-modal hallucinations, where one modality incorrectly affects predic… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Reliable Self-Evolution with Imperfect Proxy Rewards | Large language model (LLM)-based self-evolving search is a promising approach to scientific discovery. However, high-fidelity evaluati… | ↗Not briefed |
bluesky bluesky-ai | 1mo | OpenAi tech does not improve or enhance information assembled by experts and published on official websites. It churns through precious and… | OpenAi tech does not improve or enhance information assembled by experts and published on official websites. It churns through preciou… | ↗Not briefed |
arxiv arxiv-ai | 1mo | CreateScore: Domain-Theory-Informed Bayesian Routing for LLM-Based CV Screening | Large language models (LLMs) can support rubric-based screening of CVs, but applying a high-capability model to every candidate and cr… | ↗Not briefed |
arxiv arxiv-ai | 1mo | A Guideline-Augmented Multi-Agent Framework for Schema-as-Code Biomedical Named Entity Recognition | Large language models (LLMs) have shown promising potential for biomedical named entity recognition (BioNER) through instruction follo… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Audionaut – an open-source cross-platform multitrack audio editor | — | ↗Not briefed |
arxiv arxiv-ai | 1mo | Reasoning with Evidence, Not Merely Rationales: Verifiable Preference Proofs for LLM-Based Recommendation | Large language models (LLMs) can infer user preferences from interaction histories and reviews, yet the rationales they generate may n… | ↗Not briefed |
bluesky bluesky-ai | 1mo | This makes me think that Anthropic is a truly Californian company. Yes, Silicon Valley and venture capital but also cults, woo woo and vagu… | This makes me think that Anthropic is a truly Californian company. Yes, Silicon Valley and venture capital but also cults, woo woo and… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Post-Training Frontier Text-to-Image Models by Composing Preference and Rubric Rewards🔧 Ideogram | Recent text-to-image generation models have achieved remarkable visual quality, but improving them through post-training remains chall… | ↗Not briefed |
bluesky bluesky-ai | 1mo | by 'hack' it again sounds very soft: information that is public, but accessed in an unauthorised way. NSW gov and OpenAI both say no person… | by 'hack' it again sounds very soft: information that is public, but accessed in an unauthorised way. NSW gov and OpenAI both say no p… | ↗Not briefed |
bluesky bluesky-ai | 1mo | regular reminder that publicly available information is useful unless or until it is attacked by OpenAi tech, which steals data to assemble… | regular reminder that publicly available information is useful unless or until it is attacked by OpenAi tech, which steals data to ass… | ↗Not briefed |
bluesky bluesky-ai | 1mo | このイマココナビはClaudeで作ったらしいけど、Claudeは軍事AI企業パランティアのAIの基盤でもある。 それで作ったサイトに情報を入れる、ということは個人情報がClaudeやパランティアに漏洩するということになるのではないか。 | このイマココナビはClaudeで作ったらしいけど、Claudeは軍事AI企業パランティアのAIの基盤でもある。 それで作ったサイトに情報を入れる、ということは個人情報がClaudeやパランティアに漏洩するということになるのではないか。 | ↗Not briefed |
rss mit-tech-review-ai | 1mo | Don’t be fooled—LLMs don’t reason | On an afternoon in Seoul in March 2016, I watched a program I helped build put a stone on the fifth line of a Go board in what looked… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Dirac-Interconnected Neural Elements: Discovering Modularity in Physical Systems Without Reduction | Deep learning has shown remarkable success in the data-driven modeling of dynamical systems. Much of its success is attributed not to… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Ai detectors DO NOT WORK. They are based on the same LLM garbage and know as much stuff as ChatGPT does, which is to say they don't know wh…🔧 ChatGPT | Ai detectors DO NOT WORK. They are based on the same LLM garbage and know as much stuff as ChatGPT does, which is to say they don't kn… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Webdesk a Linux desktop in the browser over SSH | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | a friday afternoon surprise: NSW gov says OpenAI disclosed a "misalignment" by its AI agent and the Department of Climate Change, Energy, t… | a friday afternoon surprise: NSW gov says OpenAI disclosed a "misalignment" by its AI agent and the Department of Climate Change, Ener… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Dickhead day on Fetlife again. I’ve had over 10 messages this week of a similar vein. This one was the only plank not to disguise it at lea… | Dickhead day on Fetlife again. I’ve had over 10 messages this week of a similar vein. This one was the only plank not to disguise it a… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Understanding Trajectory Heterogeneity in Federated World Model Learning | World models learn state evolution from trajectories, making access to temporal context a central training requirement. Federated lear… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Interessant sind die Einschätzungen der beiden Gäste: Jonas Freund, Senior Research Fellow am Center for the Governance of AI. Matteo Ros…🔧 ChatGPT | Interessant sind die Einschätzungen der beiden Gäste: Jonas Freund, Senior Research Fellow am Center for the Governance of AI. Matte… | ↗Not briefed |
hn hn-ai | 1mo | Tent Shamiyana Booking PHP Script – Complete Bartan and Event Booking System | New: Tent Shamiyana Booking PHP Script – Complete Bartan & Event Booking System<p>Wedding, Birthday, Party ya kisi bhi event ke li… | ↗Not briefed |
bluesky bluesky-ai | 1mo | It seems to me that Anthropic didn't even TRY with the Muslims. I am delighted to not be represented here 😂 | It seems to me that Anthropic didn't even TRY with the Muslims. I am delighted to not be represented here 😂 | ↗Not briefed |
rss bing-news-ai | 1mo | Slovenia’s .si domain sees a surge in registrations after Trump’s ‘super intelligence’ order | The .si domain name is seeing unprecedented demand after President Trump's super intelligence executive order. | ↗Not briefed |
rss bing-news-ai | 1mo | Slovenia’s .si domain sees a surge in registrations after Trump’s ‘super intelligence’ order | The .si domain name is seeing unprecedented demand after President Trump's super intelligence executive order. | ↗Not briefed |
rss bing-news-ai | 1mo | Slovenia’s .si domain sees a surge in registrations after Trump’s ‘super intelligence’ order | The .si domain name is seeing unprecedented demand after President Trump's super intelligence executive order. | ↗Not briefed |
arxiv arxiv-ai | 1mo | Hyperparameter selection for equation learning with biologically-informed neural networks | Biologically-informed neural networks (BINNs) have emerged as a flexible subclass of physics-informed neural networks (PINNs) for lear… | ↗Not briefed |
arxiv arxiv-ai | 1mo | SlimKV: Joint Token-Feature KV Cache Compression with Reconstruction-Free Beacon Attention | Long-context LLM serving is increasingly bottlenecked by KV-cache memory, especially in resource-constrained scenarios. Among existing… | ↗Not briefed |
arxiv arxiv-ai | 1mo | GTDD: Generative Test-Driven Development for AI Coding Agents with Adversarial Testing | Test-driven development gives AI coding agents executable requirements for implementing software. Because these agents can adapt their… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Dynamic Expert Pruning for Multi-Agent Systems | Mixture-of-Experts (MoE) architectures scale language models efficiently by activating only a few experts per token, but the saving is… | ↗Not briefed |
rss the-decoder | 1mo | Black Forest Labs launches Flux 3 Image with multi-step editing that leaves the rest of your picture alone🔧 FLUX | Black Forest Labs has released Flux 3 Image, the image side of its Flux 3 model family. It supports multi-step edits without changing… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Enhancing Biomedical Named Entity Recognition via Multiple Programming Languages Instruction Tuning and Ensemble Method | Instruction tuning has become a common paradigm for applying large language models (LLMs) to biomedical named entity recognition (BioN… | ↗Not briefed |
bluesky bluesky-ai | 1mo | AI chatbots turn users’ most intimate data into a potential goldmine for advertisers, researchers warn. Six out of nine of the most popula… | AI chatbots turn users’ most intimate data into a potential goldmine for advertisers, researchers warn. Six out of nine of the most p… | ↗Not briefed |