Robot hands learned to play Nettspend from scratch with PPO! F1 reached 0.80 on the song, in range of RoboPianist and RP1M report for their per-song RL policies.
On July 25, we hacked OpenAI.
Two bugs let us take over ChatGPT/Codex accounts of OpenAI employees (+some unaffiliated users) and reach connected services: Outlook, Slack, GitHub, etc.
We proved it with a PR in OpenAIโs internal codebase . It took us <72h. ๐งต