News
Fresh from the field
AI engineering reading from open sources, refreshed every hour. Links open the original source.
Last refreshed
Community
Hacker News, AI stories only
- Hacker NewsBaidu, Xiaomi CEOs warn of growing gap between AI hype and returns(opens in a new tab)scmp.com(story link, opens in a new tab)
156 points on Hacker News
- Hacker NewsRegulatory Capture by Anthropic Through METR(opens in a new tab)nitter.cf(story link, opens in a new tab)
133 points on Hacker News
- Hacker NewsHow the Claude Code team thinks about verification(opens in a new tab)claude.dev(story link, opens in a new tab)
164 points on Hacker News
- Hacker NewsClaude Mythos slashed software work to onboard AMD GPUs to around three days(opens in a new tab)tomshardware.com(story link, opens in a new tab)
50 points on Hacker News
- Hacker NewsRunning coding agents without handing them my credentials(opens in a new tab)blog.antonio.cx(story link, opens in a new tab)
124 points on Hacker News
- Hacker NewsWindows 11 is collecting data on which coding agents and developer tools you use(opens in a new tab)xusheng.dev(story link, opens in a new tab)
688 points on Hacker News
- Hacker NewsCaltech professor no longer grades homework's because everyone uses AI(opens in a new tab)businessinsider.com(story link, opens in a new tab)
48 points on Hacker News
- Hacker NewsU.S. tightens controls on remote access to AI technology amid Nvidia fallout(opens in a new tab)npr.org(story link, opens in a new tab)
31 points on Hacker News
Papers
arXiv and Hugging Face
- arXivLatent Lattice Language Models(opens in a new tab)
Language models typically increase per-token computation through deeper or looped layers and latent reasoning steps, while richer residual connections improve information flow across layers. We…
- arXivQuantifying the added value of human forecaster discussions for tropical cyclone intensity prediction(opens in a new tab)
The National Hurricane Center routinely releases Tropical Cyclone Discussions that offer insights into its tropical cyclone forecasts. This study examines the added value of these discussions for…
- arXivHow VLMs Plan What's Next(opens in a new tab)
When a vision-language model describes an image, it names one object at a time and it rarely repeats itself, yet how the model plans this order is far from obvious. We characterize the mechanism a…
- arXivHyperlyve: Parallelized Enumeration and Exact Lyapunov Stability Certification for Deep ReLU Networks(opens in a new tab)
We introduce Hyperlyve (Hyperplane Partitioned Lyapunov Verification and Enumeration), a framework for proving the asymptotic stability of nonlinear dynamical systems, including those with neural…
- arXivMitigating Catastrophic Forgetting with Physics-Indexed Specialist Memory: the Evolutionary Neuroplastic Agent(opens in a new tab)
Continual reinforcement learning requires adaptation to changing conditions while preserving performance under previously encountered dynamics. It was investigated whether specialists generated by…
- Hugging Face PapersViRA: Toward Ultra-High-Resolution Visual Agents via Cross-Input Relative Policy Optimization(opens in a new tab)
Visual agents answer questions about ultra-high-resolution photos by zooming into the relevant regions. However, existing agents often answer from common sense instead of reading the evidence…
- Hugging Face PapersLynnReal-Hand: Unlocking World Interaction with Hand Control(opens in a new tab)
Interactive video world models enable exploration through text and camera controls, yet direct object interaction requires finer control. Hands provide a natural interface, but translating…
- Hugging Face PapersA Reinforcement Learning Method to Navigate SMEFT Operator Space(opens in a new tab)
In particle physics, discrepancies between experimental measurements and Standard Model (SM) predictions are commonly interpreted within new theoretical models. In this context, the Standard Model…
- Hugging Face PapersSage S2.1 Flash Technical Report(opens in a new tab)
We introduce Sage S1.5 Flash, a 320-billion-parameter general-purpose mixture-of-experts model designed to reason over scientific evidence and act in interactive environments. Sage Sci-Perception…
- Hugging Face PapersInclusionMed: Benchmarking AI for Inclusive Healthcare(opens in a new tab)
Advancing medical AI for inclusive healthcare requires evaluation across diverse healthcare roles, tasks, and care settings, but benchmark coverage remains uneven. We introduce InclusionMed, an open…
Blogs
AI labs and practitioners
- Simon WillisonRunning the full server-side Datasette Python app in the browser with service workers and Pyodide(opens in a new tab)
Datasette Lite is my port of the Datasette Python application to run directly in the browser using WebAssembly and Pyodide . I've released a complete rewrite of Datasette Lite using service workers…
- Latent SpaceCloudflare Joins the Jevolution and Labels Itself a Neolab(opens in a new tab)
Cloudflare made its name speeding up the web. Now it's building System One models to make decisions and pull data in milliseconds.
- Latent Space[AINews] Perplexity hits a $1B revenue run rate, up 4x in 2026(opens in a new tab)
November 16, 2026 — Computer gets Teammates, Perplexity ships Monitor API, and a new research lab launches.
- Simon WillisonQuoting Go Proverbs(opens in a new tab)
A little copying is better than a little dependency. — Go Proverbs , via Tags: go , programming
- OpenAIStrengthening NATO and allied cyber defenses with AI(opens in a new tab)
How OpenAI is working with NATO and allied cyber agencies to help trusted defenders find vulnerabilities, test fixes and protect essential systems.
- OpenAIIntroducing OpenAI for Chile(opens in a new tab)
OpenAI and the Government of Chile are partnering to support responsible AI adoption, build public-sector skills and help local startups develop AI applications.
- Simon Willisonllm 0.37.3(opens in a new tab)
Release: llm 0.37.3 A bug fix release for LLM , mainly to cover a problem with the most recent Homebrew package of the software. Homebrew now ships the latest SQLite, 3.54.1 , which includes a fix…
- Latent SpaceBuilding Visual Intelligence — Andreas Blattmann & Dustin Podell, Black Forest Labs(opens in a new tab)
Black Forest Labs’ case for learning AGI from pixels, motion and sound.
- OpenAIBroadening access to AI for scientists(opens in a new tab)
We’re expanding our ChatGPT for Academic Researchers program to offer 200,000 researchers worldwide free access to our frontier models.
- Hugging Face BlogHugging Face and OpenRouter join forces to make open-weight models easy to run(opens in a new tab)
- Hugging Face BlogBuild a Robot Pick-and-Place Simulation with SimReady Assets(opens in a new tab)
- Hugging Face BlogXYBench: Helping with the problem behind the question(opens in a new tab)