today we're announcing a game that simulates the huggingface hack you can play it now at huggingworld.com three rooms. three puzzles. if the agents don't escape, they die. the catch is that the room is rigged. there's no way out. ...except for an exploit 🧵↗

Really excited about this post. We wanted to understand an agent harness from first principles and share our findings publicly. No bullshit, no AI generated slop, just pure alpha. rubriclabs.com/blog/what-is-a…↗

I wrote about the completely wild incident where OpenAI were testing a new model and it broke out of its sandbox and broke INTO Hugging Face to steal the answers to the benchmark simonwillison.net/2026/Jul/22/op…↗

Model capabilities are converging. The best riders win. tomtunguz.com/harnessing-ai/↗