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 |
|---|---|---|---|---|
rss the-decoder | 2d | Microsoft is replacing outside AI models in its own products with cheaper in-house alternatives | Microsoft's AI division MAI unveiled seven new models for code, security, images, language, and routing at Ignite. MAI Code 2 reaches… | ↗Not briefed |
bluesky bluesky-ai | 2d | “Anthropic warns that recursively self-improving AI could concentrate power and escape human control, yet races to build it first. The comp… | “Anthropic warns that recursively self-improving AI could concentrate power and escape human control, yet races to build it first. The… | ↗Not briefed |
bluesky bluesky-ai | 2d | You don’t know how secure the code Claude or whatever generated for you is. You’re incapable of looking at the code and understanding any o… | You don’t know how secure the code Claude or whatever generated for you is. You’re incapable of looking at the code and understanding… | ↗Not briefed |
arxiv arxiv-ai | 2d | Decomposing SGD Dynamics in Neural Networks: Teacher-Induced Spikes and Variance Inflation | We study two-layer neural networks trained by stochastic gradient descent (SGD) on a multi-index teacher model, focusing on how first-… | ↗Not briefed |
bluesky bluesky-ai | 2d | Anthropic options could begin trading on Monday, expected to draw strong demand reut.rs/4zT3icZ | Anthropic options could begin trading on Monday, expected to draw strong demand reut.rs/4zT3icZ | ↗Not briefed |
bluesky bluesky-ai | 2d | This is a great example of how asking the wrong question gets you the wrong answer: www.niemanlab.org/2026/11/afte... | This is a great example of how asking the wrong question gets you the wrong answer: www.niemanlab.org/2026/11/afte... | ↗Not briefed |
hn hn-ai | 2d | My Empire of Sunshine and Roses | — | ↗Not briefed |
bluesky bluesky-ai | 2d | propos nothing, do you think it will be easier for valedictorians to claim their shoes than people claiming their anthropic settlement money | propos nothing, do you think it will be easier for valedictorians to claim their shoes than people claiming their anthropic settlement… | ↗Not briefed |
bluesky bluesky-ai | 2d | After gutting newsrooms, McClatchy plans to rebuild itself around AI buff.ly/VgtXaS8 | After gutting newsrooms, McClatchy plans to rebuild itself around AI buff.ly/VgtXaS8 | ↗Not briefed |
arxiv arxiv-ai | 2d | Equivariant Modeling with Polynomial Features via Minimal Homogeneous Generators | We present a general framework for modeling and learning equivariant functions between finite dimensional semisimple real group repres… | ↗Not briefed |
bluesky bluesky-ai | 2d | I'm getting ads for classes that promise "learn how to use Claude AI." I don't need a class on how to use AI tools for the same reason I do… | I'm getting ads for classes that promise "learn how to use Claude AI." I don't need a class on how to use AI tools for the same reason… | ↗Not briefed |
rss techmeme | 2d | Google rolls out Project Genie updates to AI Ultra subscribers, including Genie 4, new Genie 3 features, an agent to build worlds, and a gallery to share them (Google) | Google: Google rolls out Project Genie updates to AI Ultra subscribers, including Genie 4, new Genie 3 features, an agent to build wor… | ↗Not briefed |
bluesky bluesky-ai | 2d | What are we even doing here? nypost.com/2026/11/18/u... | What are we even doing here? nypost.com/2026/11/18/u... | ↗Not briefed |
hn hn-ai | 2d | Show HN: A deterministic typesetting tool for Orthodox Music | I'm a trained Russian Orthodox church musician and software developer. I figured I would flog my Claude Pro subscription into hel… | ↗Not briefed |
arxiv arxiv-ai | 2d | On Evaluating User Simulators for Interactive Agents | User simulators offer a scalable way to evaluate interactive agents, but how to assess their quality remains underexplored. In particu… | ↗Not briefed |
rss theverge-ai | 2d | What is Nvidia without AI’s sugar daddy in charge? | Building a successful company means that it can run smoothly without you. Did Jensen Huang actually build a successful company? | Imag… | ↗Not briefed |
arxiv arxiv-ai | 2d | Cap-and-Track: Optimizing LLM Inference Request Dispatch Across Shifting Capacity | A fixed dispatch rate for LLM inference becomes stale as serving fleets resize and request sizes change. Generative output length is u… | ↗Not briefed |
hn hn-ai | 2d | Show HN: Pf2e.fyi, an Obsidian-inspired searchable index of Pathfinder 2e rules | Hey folks! I was looking for a more interesting interface to search rules for the Pathfinder 2nd edition tabletop roleplaying game. As… | ↗Not briefed |
arxiv arxiv-ai | 2d | Who are you? The Personal Imitation Game🔧 Replicate | Large Language Models (LLMs) now perform strongly on Alan Turing's 1950 Imitation Game, so we propose a harder challenge, the Personal… | ↗Not briefed |
arxiv arxiv-ai | 2d | Ground, Iterate, Trace: Decomposing AI Research Ideation | Large language models have been widely used to generate plausible research ideas, yet it remains unclear what makes automated ideation… | ↗Not briefed |
hn hn-ai | 2d | Optimize Your Mwmbl Crawler | — | ↗Not briefed |
bluesky bluesky-ai | 2d | I’ve been a long time Proton user who has happily recommended their product for folk up until their CEO has gone on multiple unhinged rants… | I’ve been a long time Proton user who has happily recommended their product for folk up until their CEO has gone on multiple unhinged… | ↗Not briefed |
bluesky bluesky-ai | 2d | Instacart’s CTO stopped his engineers from reading code. 97% of their code is now AI-generated, and their shopping agent costs 1/8 what it… | Instacart’s CTO stopped his engineers from reading code. 97% of their code is now AI-generated, and their shopping agent costs 1/8 wh… | ↗Not briefed |
bluesky bluesky-ai | 2d | não posso mandar as credenciais direto pro claude mas ele precisa delas aí eu vou brincar de esconder as credenciais num arquivo p ele acha… | não posso mandar as credenciais direto pro claude mas ele precisa delas aí eu vou brincar de esconder as credenciais num arquivo p ele… | ↗Not briefed |
bluesky bluesky-ai | 2d | AI Checkers are terrible. I don't even use them to check AI for my student's work. There's a thousand other ways then feeding the LLM machi… | AI Checkers are terrible. I don't even use them to check AI for my student's work. There's a thousand other ways then feeding the LLM… | ↗Not briefed |
arxiv arxiv-ai | 2d | Monoculture Robust Learning | We introduce a notion of monoculture robustness in Probably Approximately Correct (PAC) learning, as a way to mitigate the problem of… | ↗Not briefed |
bluesky bluesky-ai | 2d | Algunos detalles que son CINEMA - La ilustración sin capas en Photoshop - No le cambiaron ni el nombre a los archivos generados con Opena… | Algunos detalles que son CINEMA - La ilustración sin capas en Photoshop - No le cambiaron ni el nombre a los archivos generados con… | ↗Not briefed |
arxiv arxiv-ai | 2d | Local AI Agents in Power Electronics Simulation: Deployment, Measurements and Costs | Agentic artificial intelligence (AI) is emerging in power electronics, where it drives the simulation loop that engineers used to clos… | ↗Not briefed |
arxiv arxiv-ai | 2d | Auditing black-box clinical prediction models with a large language model through queries alone: a blinded test with planted flaws | Large language models (LLMs) are used to explain clinical prediction models, on the promise that they can say what a model computes, f… | ↗Not briefed |
bluesky bluesky-ai | 2d | To paraphrase Rebecca Solnit, the AI industry assumes everyone is floundering with even the simplest mental activity and will welcome LLM s… | To paraphrase Rebecca Solnit, the AI industry assumes everyone is floundering with even the simplest mental activity and will welcome… | ↗Not briefed |
arxiv arxiv-ai | 2d | Learning to Interact in Multi-Agent Self-Play for Controllable Autonomous Driving Scenario Generation | Large-scale, multi-agent self-play trains robust driving policies effectively in simulation, but controlling these models to generate… | ↗Not briefed |
arxiv arxiv-ai | 2d | TangentPDE: Learning Local Operator Responses with Sparse Finite-Difference Supervision | Neural PDE operators are often optimized through state reconstruction, which largely constrains operator values at observed inputs but… | ↗Not briefed |
arxiv arxiv-ai | 2d | WorldRuler: Interactive World Modeling under Configurable Rules | Rules govern how player actions change interactive worlds. Different rules create distinct gameplay experiences within similar scenes… | ↗Not briefed |
arxiv arxiv-ai | 2d | Choosing What to Observe: Budget-Paired Counterfactual Audits for Credible Evaluation of Visual Agents | Visual agents that read dense documents must ration a finite observation budget across candidate regions, yet answer accuracy conflate… | ↗Not briefed |
bluesky bluesky-ai | 2d | Nate Soares and Beff Jezos sign the same letter accusing Anthropic of being unsafe when building RSI the list keeps going, the world agree… | Nate Soares and Beff Jezos sign the same letter accusing Anthropic of being unsafe when building RSI the list keeps going, the world… | ↗Not briefed |
bluesky bluesky-ai | 2d | 1/7 🤖 CALIFORNIA SUES OPENAI OVER 80+ AI ESCAPES 🤖 AG Rob Bonta alleges OpenAI tested models without safeguards, letting agents hack rea… | 1/7 🤖 CALIFORNIA SUES OPENAI OVER 80+ AI ESCAPES 🤖 AG Rob Bonta alleges OpenAI tested models without safeguards, letting agents hac… | ↗Not briefed |
hn hn-ai | 2d | Surveillance Isn't Safety: A Comic About Organizing Against Flock and Other Foes | — | ↗Not briefed |
bluesky bluesky-ai | 2d | “Anthropic can’t stand still...the industry could end up like Boeing and Airbus engag[ing] in price wars and neither is very profitable.” … | “Anthropic can’t stand still...the industry could end up like Boeing and Airbus engag[ing] in price wars and neither is very profitabl… | ↗Not briefed |
bluesky bluesky-ai | 2d | can claude bring back the spicy pickles? no???? then why even HAVE AI???? | can claude bring back the spicy pickles? no???? then why even HAVE AI???? | ↗Not briefed |
bluesky bluesky-ai | 2d | 📣‼️ “A terrorist can log onto an AI platform & in less than 12 hours walk away w/a concrete playbook for how to carry out a deadly attack.… | 📣‼️ “A terrorist can log onto an AI platform & in less than 12 hours walk away w/a concrete playbook for how to carry out a deadly at… | ↗Not briefed |
bluesky bluesky-ai | 2d | Wow AI models really are catching up to humans. | Wow AI models really are catching up to humans. | ↗Not briefed |
bluesky bluesky-ai | 2d | Ik dacht. “Laat ik de LLM van Canva eens een kans geven om Mock ups te genereren van dekentjes waar ik geen fysiek exemplaar heb. Is vrij … | Ik dacht. “Laat ik de LLM van Canva eens een kans geven om Mock ups te genereren van dekentjes waar ik geen fysiek exemplaar heb. Is… | ↗Not briefed |
rss deepmind-blog | 2d | Genie 4 is now available | — | ↗Not briefed |
hn hn-ai | 2d | Show HN: Re-SiON – 40kb software synthesizer library that plays plaintext files | Hi HN!<p>Re-SiON is a re-implementation of an old software synthesizer created for Adobe Flash. It's incredibly tiny (at ~40KB) a… | ↗Not briefed |
arxiv arxiv-ai | 2d | On the Impact of Feature Reduction Techniques on Adversarial Attacks | Adversarial attacks have posed significant threats to machine learning models by exploiting their inherent vulnerabilities. Researcher… | ↗Not briefed |
bluesky bluesky-ai | 2d | My LLM knows I like the crust cut off of my sandwiches... | My LLM knows I like the crust cut off of my sandwiches... | ↗Not briefed |
arxiv arxiv-ai | 2d | Can Programs Replace LLM Judges? Executable Program Judges for Scalable and Efficient Relevance Labeling | Large language model (LLM) judges have become a common way to create large-scale relevance judgments for retrieval evaluation. However… | ↗Not briefed |
bluesky bluesky-ai | 2d | Live now: Agentic Statistical Programming with Local Models @eddelbuettel.com YT: www.youtube.com/live/esBxQaH... GH: github.com/eddel… | Live now: Agentic Statistical Programming with Local Models @eddelbuettel.com YT: www.youtube.com/live/esBxQaH... GH: github.com/… | ↗Not briefed |
arxiv arxiv-ai | 2d | Atomic MathReasoner: Instance-Level Composition of Atomic Reasoning Skills for Mathematical LLMs | Most existing approaches enhance mathematical reasoning in large language models (LLMs) as a monolithic capability, despite the fact t… | ↗Not briefed |
bluesky bluesky-ai | 2d | tomorrow. also anthropic IPO. it's going to be an incredible day online | tomorrow. also anthropic IPO. it's going to be an incredible day online | ↗Not briefed |