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 | Anthropic executives thought maybe the Pope and the Catholic Church in general hadn't thought too much about it and could be convinced by t… | Anthropic executives thought maybe the Pope and the Catholic Church in general hadn't thought too much about it and could be convinced… | ↗Not briefed |
rss latent-space | 1mo | Inside-Out AI: Rebuilding Airbnb Behind the Scenes and Across the Guest Experience | After leading Meta’s Llama models, Ahmad Al-Dahle is now transforming Airbnb with AI — from how its teams develop products to how it s… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Equivariant Visual-Tactile Diffusion Policy for Contact-Rich Manipulation | Imitation learning for contact-rich manipulation requires high-quality expert data that is expensive to obtain. This makes learning a… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Game where you control a swarm | Why I thought this might be interesting for HN – there is a dynamic-code-insertion feature. When you reach 2000 units, you will get an… | ↗Not briefed |
hn hn-ai | 1mo | Local LLM Won't Stop Talking? It's Not the Model | — | ↗Not briefed |
arxiv arxiv-ai | 1mo | Cordial Learning: Distributed Training with Correlated Data | We consider a distributed learning task with agents that have correlated data. Specifically, the label of an agent depends on the inpu… | ↗Not briefed |
arxiv arxiv-ai | 1mo | SyntaxBench: A Statistical Diagnostic Framework for Character-Level Reasoning in Large Language Models | Large language models are increasingly used where small syntactic errors matter, yet character-level reasoning is still evaluated most… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Preserving Mathematical Reasoning in Compressed Diffusion Language Models via Trajectory-Aware Low-Rank Approximation | Diffusion language model (dLLM) compression faces a known challenge because calibration is typically performed on clean, fully visible… | ↗Not briefed |
rss techcrunch-ai | 1mo | Last 24 hours: Exhibit at TechCrunch Disrupt 2026 and reach 10,000+ tech leaders | Today is the very last day to book your exhibit table at TechCrunch Disrupt 2026. On October 13–15, 10,000+ founders, investors, opera… | ↗Not briefed |
arxiv arxiv-ai | 1mo | To Jev or Not? Evaluating the Accuracy and Efficiency of Structured Decision Models for Hate-Speech Moderation | The scale of online content makes hate-speech moderation challenging, while Large Language Models (LLMs) enable harmful material to be… | ↗Not briefed |
bluesky bluesky-ai | 1mo | 2030. La CSJ otorga personalidad jurídica a los sistemas de IA. 2031. La sociedad de un grupo empresario de medios y terminales de transpo… | 2030. La CSJ otorga personalidad jurídica a los sistemas de IA. 2031. La sociedad de un grupo empresario de medios y terminales de tr… | ↗Not briefed |
hn hn-ai | 1mo | Setup Diffusers for Stable Diffusion XL Using a Nvidia Pascal GPU in Linux🔧 Stable Diffusion | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | Slop mathematics =anthropic But our CA dem politicians are shivving it down our throat. @xavierbecerra.bsky.social no more @scottwiener.b… | Slop mathematics =anthropic But our CA dem politicians are shivving it down our throat. @xavierbecerra.bsky.social no more @scottwie… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Information Limits of Low-Rank Approximation Certification | Low-rank approximation can require additional matrix--vector products to verify that its error meets a prescribed tolerance. We charac… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Refinement Buys Intelligibility, Search Buys Identity: What Test-Time Compute Buys in Masked-Diffusion TTS | Diffusion language models for text-to-speech combine two forms of computation: model depth (parameters) and refinement steps (inferenc… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: ScopeLock – Autonomous Git scope-creep auditor and UCC 2-209 engine | — | ↗Not briefed |
hn hn-ai | 1mo | Gemini vs. Elevenlabs: TTS and Voice Cloning🔧 ElevenLabs | — | ↗Not briefed |
arxiv arxiv-ai | 1mo | Multi-Task Evolution for Zero-Shot Cross-Problem Generalization using LLMs | Designing effective heuristics for diverse combinatorial optimization problems requires substantial expertise and repeated search. Lar… | ↗Not briefed |
bluesky bluesky-ai | 1mo | A brilliantly maddening example of this is Meta poaching engineers from OpenAI with $100 million comp packages while USAID was gutted | A brilliantly maddening example of this is Meta poaching engineers from OpenAI with $100 million comp packages while USAID was gutted | ↗Not briefed |
arxiv arxiv-ai | 1mo | Lightweight, Rubric-Guided Trajectory Evaluation for Production AI Agents | Trajectory evaluation is essential for improving the reliability of LLM-based agents, but production use makes it expensive to run rep… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Daniel Stone examines the OpenAI agent incident in Australia: "84 days, almost three months, from the first access to the warning being rai… | Daniel Stone examines the OpenAI agent incident in Australia: "84 days, almost three months, from the first access to the warning bein… | ↗Not briefed |
bluesky bluesky-ai | 1mo | "oh you're Ed, the Anthropic Superfan" | "oh you're Ed, the Anthropic Superfan" | ↗Not briefed |
arxiv arxiv-ai | 1mo | DAWIS: Data Assimilation with Windowed Inverse Sampling via Multitask Interpolants | Flow- and diffusion-based generative models have recently emerged as flexible and highly efficient forecasting models for dynamical sy… | ↗Not briefed |
arxiv arxiv-ai | 1mo | SDECast: Probabilistic Weather Forecasting in Continuous Time with Neural SDEs | Existing machine learning weather forecasting models typically generate forecasts through autoregressive rollouts at a fixed temporal… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Optimal Planning in a Dynamic World | Background: We address the problem of planning when the set of feasible states or actions changes over time. For example, in the probl… | ↗Not briefed |
bluesky bluesky-ai | 1mo | How is nobody at OpenAI going to jail? This is baffling. I’m baffled. | How is nobody at OpenAI going to jail? This is baffling. I’m baffled. | ↗Not briefed |
bluesky bluesky-ai | 1mo | They're just another LLM. | They're just another LLM. | ↗Not briefed |
hn hn-ai | 1mo | Decision models like Jev don't beat LLM-as-a-judge or traditional classifiers | — | ↗Not briefed |
arxiv arxiv-ai | 1mo | Training-Loss Guarantees for Muon with Finite-Step Newton--Schulz Orthogonalization | Existing convergence analyses of Muon either assume exact orthogonalization or analyze classical Newton--Schulz polynomials, and guara… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: SOC2 Auditor Skill | SOC2 isn't affordable for early bootstrapped startups but some clients still want some way to verify trust claims.<p>I created a… | ↗Not briefed |
rss techmeme | 1mo | Sources: an EU investigation is set to find that Azure and AWS meet DMA thresholds, as the EU strengthens its regulatory scrutiny of the cloud sector (Bloomberg) | Bloomberg: Sources: an EU investigation is set to find that Azure and AWS meet DMA thresholds, as the EU strengthens its regulatory sc… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Si vous commencez le vibe coding, faites tout en Rust. Le compilateur est chiant, contraignant, il vous dit exactement ce qui ne va pas. Et… | Si vous commencez le vibe coding, faites tout en Rust. Le compilateur est chiant, contraignant, il vous dit exactement ce qui ne va pa… | ↗Not briefed |
arxiv arxiv-ai | 1mo | S$^{2}$-PINN: Stochastic Separable Physics-Informed Neural Networks | Uncertainty quantification (UQ) for random partial differential equations (PDEs) is ubiquitous in computational science and engineerin… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Falvie and Loish should get a billion dollars from OpenAI | Falvie and Loish should get a billion dollars from OpenAI | ↗Not briefed |
hn hn-ai | 1mo | Extracting Tasks from Millions of Agent Runs | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | Why it look AI generated? I don't think it's the case. They, at least, seems to have real actors and a minimal set. Somehow the image and t… | Why it look AI generated? I don't think it's the case. They, at least, seems to have real actors and a minimal set. Somehow the image… | ↗Not briefed |
arxiv arxiv-ai | 1mo | JOVE: Joint Execution and Verification for Resource-Aware LLM Task Graphs | Complex reasoning queries can be decomposed into directed acyclic task graphs and distributed across heterogeneous LLMs, reducing late… | ↗Not briefed |
bluesky bluesky-ai | 1mo | This is insane. | This is insane. | ↗Not briefed |
arxiv arxiv-ai | 1mo | Wrong Organ, Right Physics: Transferring Echocardiography Pretraining to Lung Ultrasound for Tuberculosis Screening | Lung ultrasound (LUS) is attractive for tuberculosis (TB) screening at primary-care level, but labelled cohorts are small. Echocardiog… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Architecture-Dependent Fusion Pathways in MLLMs🔧 Modal | Multimodal Large Language Models (MLLMs) achieve strong performance across vision-language tasks, yet the internal mechanisms by which… | ↗Not briefed |
hn hn-ai | 1mo | Nvidia/OpenShell: Safe, private runtime for autonomous AI agents | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | Having one of those mornings where I'm wondering "is LLM engineering really all that much better?" The complexity creep is so hard to figh… | Having one of those mornings where I'm wondering "is LLM engineering really all that much better?" The complexity creep is so hard to… | ↗Not briefed |
bluesky bluesky-ai | 1mo | This week's Seriously Risky Business looks at OpenAI agents "infiltrating" an Australian government website, the hack of the FBI, Ukraine g… | This week's Seriously Risky Business looks at OpenAI agents "infiltrating" an Australian government website, the hack of the FBI, Ukra… | ↗Not briefed |
bluesky bluesky-ai | 1mo | NEW: OpenAI's AI agents BYPASSED SECURITY at 100+ organizations without permission. This happened during testing. If that's the controlled… | NEW: OpenAI's AI agents BYPASSED SECURITY at 100+ organizations without permission. This happened during testing. If that's the contr… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Very positive discussion with my third years about LLMs today. We talked environmental costs, authoritarianism, patriarchy, tech-bros, etc.… | Very positive discussion with my third years about LLMs today. We talked environmental costs, authoritarianism, patriarchy, tech-bros,… | ↗Not briefed |
bluesky bluesky-ai | 1mo | AI-generated text about AI’s superpowers always has a through-the-looking-glass quality. “AI can sit inside that loop, and that is genuin… | AI-generated text about AI’s superpowers always has a through-the-looking-glass quality. “AI can sit inside that loop, and that is g… | ↗Not briefed |
bluesky bluesky-ai | 1mo | "Officials from companies including OpenAI and Anthropic said they will testify – but every company excluding Meta agreed to do so only aft… | "Officials from companies including OpenAI and Anthropic said they will testify – but every company excluding Meta agreed to do so onl… | ↗Not briefed |
bluesky bluesky-ai | 1mo | fortune.com/2026/10/01/y... | fortune.com/2026/10/01/y... | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Spens sandboxed, observable coding agents | Spens is a tool created to standardized sandboxing (good enough) and observability around coding agents. It leverages docker, nono and… | ↗Not briefed |
bluesky bluesky-ai | 1mo | NEW: According to a bombshell report in the New York Times, Anthropic co-founder Chris Olah threatened to walk out of Pope Leo XIV’s AI enc… | NEW: According to a bombshell report in the New York Times, Anthropic co-founder Chris Olah threatened to walk out of Pope Leo XIV’s A… | ↗Not briefed |