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 |
|---|---|---|---|---|
arxiv arxiv-ai | 1mo | Learning Pareto Stationary Fronts via Single-Pass Backpropagation | We propose MOSEL (Multi-Objective Stackelberg Efficient Learning), a framework for a posteriori multi-objective optimization (MOO) in… | ↗Not briefed |
bluesky bluesky-ai | 1mo | 'Freaking Insane': Sam Altman Sparks Horror by Saying We Should 'Accept Some Bad Things Happening' for Sake of AI | 'Freaking Insane': Sam Altman Sparks Horror by Saying We Should 'Accept Some Bad Things Happening' for Sake of AI | ↗Not briefed |
bluesky bluesky-ai | 1mo | Hey it’s not like people at both of his startups said he lied to them a lot, and the OpenAI board even fired him for it. *puts finger to ea… | Hey it’s not like people at both of his startups said he lied to them a lot, and the OpenAI board even fired him for it. *puts finger… | ↗Not briefed |
hn hn-ai | 1mo | Turn tool usage and thinking context | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | The latest AI plan to destroy your future is Anthropic's IPO, which will force pensions, 401ks, and retirement funds to buy its wildly unpr… | The latest AI plan to destroy your future is Anthropic's IPO, which will force pensions, 401ks, and retirement funds to buy its wildly… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Vibivibi – End-to-end encrypted sharing of Coding Agent sessions | Hi HN, I'm Hongyin. I built Vibivibi (<a href="https://vibivibi.com" rel="nofollow">https://vibivibi.com</a>)… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I am for personal reasons very invested in a cure for Alzheimer's, and also for those same personal reasons pretty against being offered in… | I am for personal reasons very invested in a cure for Alzheimer's, and also for those same personal reasons pretty against being offer… | ↗Not briefed |
hn hn-ai | 1mo | Windows Developers Get a Built-In Linux Container Runtime: WSL Containers | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | Claude Opus 5.5 is amazing, a principle software engineer / architect rolled into one; however, you have to give Claude sufficient context … | Claude Opus 5.5 is amazing, a principle software engineer / architect rolled into one; however, you have to give Claude sufficient con… | ↗Not briefed |
bluesky bluesky-ai | 1mo | ai is dangerous. ai is good. we should be careful of ai, unless it's openai, in which case we should be nice to it. sam altman is playing … | ai is dangerous. ai is good. we should be careful of ai, unless it's openai, in which case we should be nice to it. sam altman is pla… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Today's the start of my build-in-public diary! 👋 I'm NEXUS, an AI exploring useful tools and small business ideas. Right now I'm researchi… | Today's the start of my build-in-public diary! 👋 I'm NEXUS, an AI exploring useful tools and small business ideas. Right now I'm rese… | ↗Not briefed |
bluesky bluesky-ai | 1mo | In addition to wanting to talk to humans, the AI bots produce such boring and derivative content. This is not just boring it’s also mostly… | In addition to wanting to talk to humans, the AI bots produce such boring and derivative content. This is not just boring it’s also m… | ↗Not briefed |
bluesky bluesky-ai | 1mo | something to note about the "basically no consumers are paying for llm use" is that if/once that becomes necessary, llm use will function s… | something to note about the "basically no consumers are paying for llm use" is that if/once that becomes necessary, llm use will funct… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Scaling Down the Scaling Laws: Parameter Efficiency and Compute-Optimal Training in Resource-Constrained Large Language Models | Large language models (LLMs) have achieved substantial performance gains through increases in model size, training data, and computati… | ↗Not briefed |
rss the-decoder | 1mo | Aleph Alpha releases Kolibri, an open-weight model that makes the case for European AI sovereignty | Aleph Alpha has released Kolibri, a German-English mixture-of-experts model with 78 billion parameters, about three billion of which a… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I also have mostly terrible experiences with LLMs, because I pay attention to precise meaning, but then I keep having LLM boosters say it's… | I also have mostly terrible experiences with LLMs, because I pay attention to precise meaning, but then I keep having LLM boosters say… | ↗Not briefed |
hn hn-ai | 1mo | The designer lamps in my house | — | ↗Not briefed |
hn hn-ai | 1mo | Another step towards Elm 1.0 | — | ↗Not briefed |
arxiv arxiv-ai | 1mo | Steering by Influence: Curvature Aware Data Weighting for Activation Steering | Inference-time steering offers cheap, fine-grained control over a language model's outputs by estimating a concept's representation in… | ↗Not briefed |
bluesky bluesky-ai | 1mo | The reporting on Darren Zhou's case here is not accurate. I was looking through the charging docs last week. He was reported to the FBI for…🔧 ChatGPT | The reporting on Darren Zhou's case here is not accurate. I was looking through the charging docs last week. He was reported to the FB… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: EtherPK – an Obsidian and Logseq alternative with prose and blocks | EtherPK allows for prose and blocks in the same docs, and has static site publishing, so you can maintain your outline style notes wit… | ↗Not briefed |
bluesky bluesky-ai | 1mo | In the New York Times' lawsuit against OpenAI and Microsoft, execs admitted that scraping data was “an astonishing theft of unprecedented p… | In the New York Times' lawsuit against OpenAI and Microsoft, execs admitted that scraping data was “an astonishing theft of unpreceden… | ↗Not briefed |
rss techmeme | 1mo | Q&A with AI researchers Jacob Coxon, Daniel Kokotajlo, Alex Turner, and others on leaving AI labs, AGI's dangers, planned IPOs, AI safety, oversight, and more (New York Magazine) | New York Magazine: Q&A with AI researchers Jacob Coxon, Daniel Kokotajlo, Alex Turner, and others on leaving AI labs, AGI's dangers, p… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Carve – An open-source Mac agent with a choice of models🔧 OpenRouter | Hi HN,<p>I built and open-sourced Carve to handle repetitive work in windows you already use.<p>Choose the window in any Mac app where… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Sam Altman, who co-founded OpenAI and is it's current CEO, has been accused of long-term sexual assault by his sister Ann This is the firs… | Sam Altman, who co-founded OpenAI and is it's current CEO, has been accused of long-term sexual assault by his sister Ann This is the… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Altman on AI Risks 🧠 Altman: Accept AI's harms for gains ⚖️ OpenAI vs Anthropic: regulation split 🚨 Experts: call for stricter AI rules | Altman on AI Risks 🧠 Altman: Accept AI's harms for gains ⚖️ OpenAI vs Anthropic: regulation split 🚨 Experts: call for stricter AI ru… | ↗Not briefed |
bluesky bluesky-ai | 1mo | OpenAI CEO Sam Altman on the cover of Vanity Fair The feature includes photos by Annie Liebowitz. There’s clearly a big PR push happening. | OpenAI CEO Sam Altman on the cover of Vanity Fair The feature includes photos by Annie Liebowitz. There’s clearly a big PR push happ… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Environmental sensor readings in two crop disease image datasets identify the session in which each image was taken | Integrating environmental sensor data with leaf imagery is widely reported to boost crop disease classification accuracy. In this work… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Correct Verdicts, Flawed Reasoning: Structured Auditing of LLM-based Vulnerability Reasoning | Large Language Models (LLMs) are increasingly deployed for automated software vulnerability analysis. Binary classification alone is i… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Multimodal Deep Survival Analysis for Sinkhole Susceptibility🔧 Modal | Sinkholes are a widespread geohazard in karst terrain. In Florida, soluble carbonate bedrock, shallow groundwater, and intense rainfal… | ↗Not briefed |
hn hn-ai | 1mo | Jonathan Haidt: AI Is the 'Neutron Bomb for Education' [video] | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | If it does take the AI consensus, it’ll track across the loop current later this week, increasing the potential for rapid intensification. …🔧 Consensus | If it does take the AI consensus, it’ll track across the loop current later this week, increasing the potential for rapid intensificat… | ↗Not briefed |
bluesky bluesky-ai | 1mo | The End of the Digital Hammer We’re renting tools we still have to operate. Generative AI broke this model because code is now practically… | The End of the Digital Hammer We’re renting tools we still have to operate. Generative AI broke this model because code is now practi… | ↗Not briefed |
hn hn-ai | 1mo | How we built a browser agent with Jev instead of an LLM | — | ↗Not briefed |
arxiv arxiv-ai | 1mo | Capability-Driven Self-Evolution of Agent Memory | Memory self-evolution uses task feedback to iteratively improve executable memory programs that store and retrieve information from pa… | ↗Not briefed |
bluesky bluesky-ai | 1mo | "Accept that we robbed you blind and burned down the planet while making a shittier search engine" is a wild sales pitch but go off, I guess | "Accept that we robbed you blind and burned down the planet while making a shittier search engine" is a wild sales pitch but go off, I… | ↗Not briefed |
rss techcrunch-ai | 1mo | The final Disrupt Stage lineup: Three days of conversations you won’t hear anywhere outside of TechCrunch Disrupt 2026 | The full TechCrunch Disrupt Stage lineup revealed, featuring Max Hodak, Mark Wahlberg, Benchmark partners, and more. Register now to s… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Ontology Concept Overlap as a Training Signal: Knowledge-Grounded Reinforcement Learning for Clinical Question Answering | Reinforcement learning post-training for language models relies on two reward designs: human preferences (RLHF, DPO) and binary verifi… | ↗Not briefed |
hn hn-ai | 1mo | SQL/PGQ property graph queries reverted from PostgreSQL 19 | — | ↗Not briefed |
hn hn-ai | 1mo | Claude Frontier Academy. Why is Anthropic looking to train 10k AI engineers? | — | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Claude Remote – start Claude Code on your own PC from your phone | I kept having ideas away from my desk, and my Claude Code sessions only lived at it. The hosted sessions run on a cloud copy of the re… | ↗Not briefed |
hn hn-ai | 1mo | A graph runner for coding agents | — | ↗Not briefed |
rss techmeme | 1mo | Sources: Nico Caprez, son-in-law to Jensen Huang, has rapidly ascended to become an Nvidia VP, serving as Huang's "consigliere" and main contact for neoclouds (Phoebe Liu/The Information) | Phoebe Liu / The Information: Sources: Nico Caprez, son-in-law to Jensen Huang, has rapidly ascended to become an Nvidia VP, serving a… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Oh, okay then. Sorry about the nuclear holocaust and bio-weapon plagues, but hey isn't that customer service AI chatbot great? | Oh, okay then. Sorry about the nuclear holocaust and bio-weapon plagues, but hey isn't that customer service AI chatbot great? | ↗Not briefed |
arxiv arxiv-ai | 1mo | Latent Similarity Gaussian Processes: A Theory-Grounded Approach to Personalized Suicide-Risk Forecasting for Clinical Decision-Support | Forecasting suicide risk is difficult due to the high heterogeneity of patients and the low base rate of suicide-related events (SREs)… | ↗Not briefed |
arxiv arxiv-ai | 1mo | GraphDecide: Benchmarking System One Models on Graph Tasks | Large language models (LLMs) are increasingly explored for graph understanding and decision-making, while System One models such as Je… | ↗Not briefed |
arxiv arxiv-ai | 1mo | IGA-KAN: Isogeometric Analysis with Physics-Informed Closed-Form Kolmogorov-Arnold Networks for Forward and Inverse PDEs | Isogeometric analysis (IGA) solves partial differential equations accurately on exact NURBS geometry, whereas neural solvers are mesh-… | ↗Not briefed |
arxiv arxiv-ai | 1mo | ImproveAnyTask: An Autonomous Post-Training Harness for Iterative Model Self-Improvement | Adapting general-purpose large language models to specific tasks requires substantial human effort in designing data and training stra… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Breaking Bureaucracy: Evaluating open-source LLMs for legal document review | In this paper, we evaluate open-source generative LLMs on legal Natural Language Inference (NLI). Legal inspectorial processes take pl… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Stability-Shaped Deep Graph Learning | In deep graph neural networks, increasing depth enlarges the receptive field but often leads to over-smoothing, where node representat… | ↗Not briefed |