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 | 16d | 👀 How to say you do not understand Ethics nor LLMs without saying you do not understand Ethics nor LLMs | 👀 How to say you do not understand Ethics nor LLMs without saying you do not understand Ethics nor LLMs | ↗Not briefed |
hn hn-ai | 16d | OpenAI's report on GPT-6 Astra underperforming on safety tasks, number by number | — | ↗Not briefed |
rss techcrunch-ai | 16d | Meta’s warns that election backlash against data centers could make AI expansion harder | Meta warns that growing political opposition to data centers could threaten future AI investments, as communities push back against ma… | ↗Not briefed |
bluesky bluesky-ai | 16d | I think my feelings would be hurt because even though I have worked *very hard* to not need everyone to like me, I still want people to lik… | I think my feelings would be hurt because even though I have worked *very hard* to not need everyone to like me, I still want people t… | ↗Not briefed |
hn hn-ai | 16d | Ask HN: How to contact Google support after cancel a subscription? | I understand this is on me for "trusting big G." Even after I've de-Googled my life as much as I could, I wanted to giv… | ↗Not briefed |
hn hn-ai | 16d | Anthropic engineer warns about 'cognitive surrender' | — | ↗Not briefed |
arxiv arxiv-ai | 16d | s$^3$: Simple Scaling of Swarms via Cooperative Multi-agent RL | Groups can achieve more than individuals in the same amount of time, but only if their members coordinate: they must decide who works… | ↗Not briefed |
bluesky bluesky-ai | 16d | OpenAI released a new cybersecurity model. Is that enough? | OpenAI released a new cybersecurity model. Is that enough? | ↗Not briefed |
bluesky bluesky-ai | 16d | ‘The anonymous seller is taking offers for a dozen loans via a so-called bids-wanted-in-competition process, …loans for sale include a $25 … | ‘The anonymous seller is taking offers for a dozen loans via a so-called bids-wanted-in-competition process, …loans for sale include a… | ↗Not briefed |
rss techcrunch-ai | 16d | OpenAI released a new cybersecurity model. Is that enough? | OpenAI's newest cybersecurity model comes out as AI labs face scrutiny of their work to protect digital infrastructure. | ↗Not briefed |
arxiv arxiv-ai | 16d | NormGym: Evaluating Normative Competence through Social Interaction | Normative competence, the ability to infer socially enforced expectations from interaction and adjust behavior accordingly, is essenti… | ↗Not briefed |
rss the-decoder | 16d | Google's Gemini Omni 1.2 Flash stretches AI video clips to 15 seconds and cuts pricing by 30 percent | Google Deepmind is making Gemini Omni 1.2 Flash generally available. The video-generation model now creates clips up to 15 seconds lon… | ↗Not briefed |
hn hn-ai | 16d | Show HN: Sshole-audit – Find, prove and fix *sshole: the OpenSSH password bypass | — | ↗Not briefed |
rss theverge-ai | 16d | Muse has a voice mode now | Meta now lets you have voice conversations with your Muse AI agent in real time, CEO Mark Zuckerberg announced today. The feature has… | ↗Not briefed |
rss techmeme | 16d | METR: Anthropic's models are improving about twice as fast as they did between March 2024 and October 2025, with internal AI use driving ~80% of acceleration (Tiffany Li/The Information) | Tiffany Li / The Information: METR: Anthropic's models are improving about twice as fast as they did between March 2024 and October 20… | ↗Not briefed |
arxiv arxiv-ai | 16d | AutoPYara: Next-Gen YARA Rule Generator for Malware Family Clustering | Millions of malware variants threaten users daily, such that clustering these samples is key for fast incident response. AutoYara is t… | ↗Not briefed |
arxiv arxiv-ai | 16d | Just CaloCHILL: fast hadronic shower generation in a learned latent space | Calorimeter showers are the most expensive part of detector simulation, and highly granular calorimeters, in which a single hadronic s… | ↗Not briefed |
bluesky bluesky-ai | 16d | A mystery seller, per Bloomberg, is trying to dump $500m in data center loans, including ones being CyrusOne (Google), Vantage Data Centers… | A mystery seller, per Bloomberg, is trying to dump $500m in data center loans, including ones being CyrusOne (Google), Vantage Data Ce… | ↗Not briefed |
bluesky bluesky-ai | 16d | The tech might've crippled the economy, but the end results speak for themselves | The tech might've crippled the economy, but the end results speak for themselves | ↗Not briefed |
arxiv arxiv-ai | 16d | User Embeddings are Superpositions of Interpretable Behavioral Modes | User embeddings are central components of modern recommendation, personalization, and behavioral modeling systems. Though these embedd… | ↗Not briefed |
arxiv arxiv-ai | 16d | SwarmEvolve: Latent Agent Swarms for Evolutionary Discovery and Optimization | LLM-guided evolutionary search uses AI agents to improve programs for mathematical and computational problems by iteratively refining… | ↗Not briefed |
arxiv arxiv-ai | 16d | HaNDF: Object-Conditioned Geometric Neural Hand Distance Fields | We propose, HaNDF, object-conditional neural distance fields as data-driven priors for modeling the plausible hand-object configuratio… | ↗Not briefed |
arxiv arxiv-ai | 16d | Scientist World: Grow a Research World for AI-Scientists | AI-Scientists arrived, but a world for them to grow in did not. Human scientists take theirs for granted $-$ centuries of open problem… | ↗Not briefed |
hn hn-ai | 16d | Show HN: Ski Voice-First Multilingual Neutral Coding Harness | Glad to showcase SKI (<a href="https://justski.it" rel="nofollow">https://justski.it</a>) , a voice-first coding h… | ↗Not briefed |
arxiv arxiv-ai | 16d | XStack: A Benchmark for ML Repo Migration🔧 GPT-5 | ML code should run where researchers choose, not where it was first written. Today, machine learning code mainly lives in two ecosyste… | ↗Not briefed |
arxiv arxiv-ai | 16d | Same Behavior, Different Mechanism: What Finite Experiments Can Certify About Integrated Information | Interpretability, AI auditing and assessments of machine consciousness increasingly rest on mechanism-level quantities: functionals of… | ↗Not briefed |
arxiv arxiv-ai | 16d | Towards interpreting any residual stream direction by scaling activation inversion | To interpret a direction in a language model's residual stream, we can look for text that strongly activates it. But when such text is… | ↗Not briefed |
bluesky bluesky-ai | 16d | Its so crazy to me that OpenAI more or less built the psychosis demon core and just left it out for thenpublic for over a year | Its so crazy to me that OpenAI more or less built the psychosis demon core and just left it out for thenpublic for over a year | ↗Not briefed |
bluesky bluesky-ai | 16d | So, uh, about that Anthropic IPO... | So, uh, about that Anthropic IPO... | ↗Not briefed |
bluesky bluesky-ai | 16d | atmospheric laptop on a Twilio x OpenAI video lfg! but seriously, this was one of the best shoots i've been on and that's in large part to… | atmospheric laptop on a Twilio x OpenAI video lfg! but seriously, this was one of the best shoots i've been on and that's in large pa… | ↗Not briefed |
bluesky bluesky-ai | 16d | Isn't Anthropic supposed to fucking PAY us soon??? THINK OF THE INTEREST ACCRUING ON THAT FUCKING PAYOUT | Isn't Anthropic supposed to fucking PAY us soon??? THINK OF THE INTEREST ACCRUING ON THAT FUCKING PAYOUT | ↗Not briefed |
bluesky bluesky-ai | 16d | openai.com/index/settin... some of this initiative was presumably already underway but I don't think you can prove it's a coincidence it wa… | openai.com/index/settin... some of this initiative was presumably already underway but I don't think you can prove it's a coincidence… | ↗Not briefed |
bluesky bluesky-ai | 16d | Dual-expert AI nails Twitter sentiment ✈️ A new dual-expert AI model achieved 98.5% accuracy on imbalanced Twitter sentiment data using tw… | Dual-expert AI nails Twitter sentiment ✈️ A new dual-expert AI model achieved 98.5% accuracy on imbalanced Twitter sentiment data usi… | ↗Not briefed |
bluesky bluesky-ai | 16d | En RTVE llevan muuuucho tiempo usando IAgen en diferentes programas y cuando nos hemos quejado a la defensora de la audiencia nos dicen que… | En RTVE llevan muuuucho tiempo usando IAgen en diferentes programas y cuando nos hemos quejado a la defensora de la audiencia nos dice… | ↗Not briefed |
rss techmeme | 16d | California's Proposition 40, which proposes a one-time 5% tax on assets of billionaires for healthcare funding, seems headed for defeat as vote counts continue (Los Angeles Times) | Los Angeles Times: California's Proposition 40, which proposes a one-time 5% tax on assets of billionaires for healthcare funding, see… | ↗Not briefed |
arxiv arxiv-ai | 16d | Adversarial Training with Large Step Sizes: Implicit Bias, Optimization, and Sharpness | Adversarial Training (AT) is an important method for improving model robustness. We study the dynamics of deterministic Gradient Desce… | ↗Not briefed |
arxiv arxiv-ai | 16d | Unpacking the Compute Efficiency of Strong-to-Weak OPD against GRPO | On-policy distillation (OPD) trains a student on its own rollouts under a frozen teacher's per-token supervision. Does this dense sign… | ↗Not briefed |
arxiv arxiv-ai | 16d | ReproBridge: Building Reproducibility into ML Research Artifacts | Reproducing machine learning experiments requires reconstructing procedures from information scattered across papers, released code, a… | ↗Not briefed |
hn hn-ai | 16d | KLM sees 1,000 fewer jobs in push to restore profitability | — | ↗Not briefed |
arxiv arxiv-ai | 16d | Q-OIL: Optimistic Exploration in Learning from Human Interventions | As robots are deployed in the real world, they will inevitably make mistakes. Human interventions provide valuable feedback for improv… | ↗Not briefed |
bluesky bluesky-ai | 16d | As I nibble kanelbulla in the Finnish Consulate, storm clouds gather over Wall Street… was this a correction or a start of a crash… will pe… | As I nibble kanelbulla in the Finnish Consulate, storm clouds gather over Wall Street… was this a correction or a start of a crash… wi… | ↗Not briefed |
bluesky bluesky-ai | 16d | "Okada's somehow escaped the cage as well!" Lads, the door in front of you was busted wide open by Wardlow. You have the object permanenc… | "Okada's somehow escaped the cage as well!" Lads, the door in front of you was busted wide open by Wardlow. You have the object perm… | ↗Not briefed |
bluesky bluesky-ai | 16d | The FTC's investigation into OpenAI & Anthropic is a start of potential accountability. But investigators need to get this right. They ca… | The FTC's investigation into OpenAI & Anthropic is a start of potential accountability. But investigators need to get this right. Th… | ↗Not briefed |
bluesky bluesky-ai | 16d | For instance, look at Nvidia, who has benefitted greatly from selling chips to companies like Anthropic and OpenAI, and invested in many di… | For instance, look at Nvidia, who has benefitted greatly from selling chips to companies like Anthropic and OpenAI, and invested in ma… | ↗Not briefed |
bluesky bluesky-ai | 16d | Who wants an applied semiotics lesson in LLM guardrails today? I'm feeling good. | Who wants an applied semiotics lesson in LLM guardrails today? I'm feeling good. | ↗Not briefed |
bluesky bluesky-ai | 16d | Husband told his wife to use ChatGPT and then it helped her plan to kill him is too good to be true. The entire AI fanbase is loser white …🔧 ChatGPT | Husband told his wife to use ChatGPT and then it helped her plan to kill him is too good to be true. The entire AI fanbase is loser w… | ↗Not briefed |
arxiv arxiv-ai | 16d | A Hidden Markov Preference Model for Online Prediction of Adversary Actions | A defender anticipating an attacker needs a probability distribution over the attacker's next action from information available before… | ↗Not briefed |
hn hn-ai | 16d | Show HN: Is Android ready for AI agents? What broke in 3324 actions on 45 phones | — | ↗Not briefed |
bluesky bluesky-ai | 16d | With access to your bank account via Plaid, Anthropic's chatbot will try to explain where your money is going and what you can afford. www… | With access to your bank account via Plaid, Anthropic's chatbot will try to explain where your money is going and what you can afford.… | ↗Not briefed |
bluesky bluesky-ai | 16d | Right? Based on the unrealistic valuations of OpenAI, Anthropic, et al? There isn’t enough money **in the world** to cover their debts. L… | Right? Based on the unrealistic valuations of OpenAI, Anthropic, et al? There isn’t enough money **in the world** to cover their debts… | ↗Not briefed |