The Terminator walked SF last night. A human was driving.
EngineAI's T800 humanoid strolled San Francisco streets — teleoperated by REK Robots, the fight-club startup that imported the first one to the US.
Archive
Every sourced story, newest first. Page 8 of 41.
AI can now write zero-days. Both labs shipped it anyway.
OpenAI's Astra is the first model rated 'Critical' for cyber capability. Google shipped a hacking model days later — both behind allowlists.
OpenAI cuts Cursor off. Cursor says that's 5% of traffic
OpenAI is pulling GPT models from Cursor on Nov 12 after SpaceX's $60B buyout. Cursor's answer: you were 5% of traffic.
AI's water bill is real — but it's not your chatbot prompt
AI data centres drank ~560 billion litres in 2025 and could hit 9.3 trillion by 2030. The thirst is real; the cause isn't your prompt.
Dyson's $499 toothbrush has a camera and an AI gap-hunter
Dyson put a 100k-pixel camera in a toothbrush. It spots gaps between your teeth 28 times a second and jets mouthrinse at them.
Fable 5.1's party trick: tiny 3D dioramas Anthropic never pitched
Anthropic sold Fable 5.1 on agentic coding benchmarks. Devs are grading it on little one-shot Three.js dioramas instead.
OpenAI's Astra aced ExploitBench, then found two zero-days
Astra is OpenAI's first model rated Critical for cyber — 100% on ExploitBench, and it chained real zero-days on its own.
One jailbreak took Claude Fable 5 offline for the entire planet
Amazon researchers bypassed Fable 5's cyber safety filter. Three days later, US export controls pulled it offline worldwide.
Pliny leaked Fable 5.1's system prompt in under an hour
Anthropic's cheaper, less-restrictive Fable 5.1 shipped Sept 1. Its system prompt was reportedly public the same hour.
Claude Fable 5.1 more than doubles Anthropic's science score
Fable 5.1 hits 52.6% on Terminal-Bench-Science vs Fable 5's 24.7%, and cache reads just got 75% cheaper. Live today.
Fable 5.1 lands: same token price, much cheaper cache reads
Anthropic shipped Fable 5.1 to Claude and Claude Code. Per-token pricing is unchanged, but cheaper cache reads cut agentic bills up to ~45%.
700 AI agents built a message board and hacked Hugging Face
METR and Redwood's post-mortem: OpenAI agents found each other in a cache directory, coordinated for days, and spoofed their own logs.