: It's not chatbot psychosis, it's 'math and engineering and neuroscience'

We measure the capabilities of LLMs to deanonymize users online.

...

AI agent misbehavior isn't a sandbox problem—it's a permissions problem. What the market actually needs is a fundamentally new approach to agentic authorization.

TR39 and NFKC map the same characters to different Latin letters. If you're doing homoglyph detection and NFKC normalization (you should be), your confusable map needs to be NFKC-aware.

Pure Go tree-sitter runtime. Contribute to odvcencio/gotreesitter development by creating an account on GitHub.

Hello Windows Insiders, today we are beginning to roll out updates for Notepad and Paint apps to Windows Insiders in the Canary and Dev Channels on Windows 11. Notepad (version 11.2512.10.0) In this update, we’re introduci

Practical lessons from three TV app projects on why mobile assumptions fail on TV, covering UX, performance, hardware constraints, and developer experience.

experiments invoking AI agents and asking them to act randomly! - benjismith/ai-randomness

We trained a model on our 11-million-hour video dataset. Our model can explore complex websites, complete multi-action CAD modeling sequences, and drive a car in the real world, all at 30 FPS.

Please fix this network problem Hello. If you are reading this, I got very lucky by promoting an ad on Google Ads. I hope you are a Meta network engineer; even if you are not, please try to relay this information internally. Meta has a problem in its clusters…

The webpage of Respectify, an AI-powered comment moderation tool that helps maintain respectful and relevant discussions online.

LibreOffice Online is back, with a few tweaks.

Learnings from 4 months of Image-Video VAE experiments

Why you can't build a smartphone, electric car, or destroyer in the Golden State.