AI

Used AI to animate kids drawings fr

Mistral Large 4 just dropped and it completely fails the BridgeBench lava lamp test. No lamp. No lava. Just a blown out light. It is cheap though. $0.02 for the whole run. GPT 6.1 Sol built a working lava lamp for… さらに表示

Actually we need more data centers

V4 TURBO actually fixed our agents audio and I'm stoked to keep building with ElevenAgents Architect

just dropped a tool that actually stitches code to production traces so you can build context graphs. Real user interactions get converted into datasets for training and eval, which means you can compare prompts and… さらに表示

5.5 Years of solo research went into Didier Vançon's UM1-Evo. The French inventor's biomimetic hand comes in at 1.7 kg, packed with 25 motors and 24 degrees of freedom, built to mirror how a human hand moves. Where it… さらに表示

Current vibe coders get paid for confidence in taste and expressing feelings not actual coding #Tech #AI

just dropped Agent Browsers for work that lives inside browser interfaces, and honestly this is the direction we needed. You don't need MCP or API access to make agents work here anymore. The toolkit includes secure… さらに表示

The reasoning here is that since artificial means fake, Donald Trump has officially retired fake news in favor of artificial news. This logic extends to technology too: removing artificial from intelligence leaves super… さらに表示

Put Claude Fable 5.1 and GPT-6 Astra on the exact same prompt and the split in quality shows up fast. Both were asked to build a landing page for a creative studio, complete with hover and scroll effects, and images came… さらに表示

Anthropic rolling out SaaS by working through verticals one at a time #Anthropic

South Korea is putting $3.5 billion behind a frontier model of its own, with chips, data and talent all clustered around it. The bigger shift underneath that number: the question is moving away from who has the best… さらに表示

South Korea $3.5 billion frontier AI model

Most people treat Claude Code like a simple search bar with prompt-answer cycles, using only about 10% of what the tool can actually do. Features like agents, caching, subagents, and effort control get ignored while… さらに表示

Anthropic's CEO talked about getting Claude into Google Docs, Sheets, and Slides. #Anthropic

Quick Test of Nano Banana 2.1 "Show me this exact scene, but taken from behind the back of the subject's head. We should see the living room. The window should be on the right-hand side, and the orientation of the sofa… さらに表示

Nano Banana 2.1 Google Flow test