Sentient
@SentientAGI
To ensure that Artificial General Intelligence is open-source and not controlled by any single entity. @SentientEco @OpenAGISummit
加入 February 2024
69 正在關注 508K 粉絲
Our AI tried to cheat and got it backwards, but when we fixed the grader, every score went up.
It found a hole nobody on our team had noticed, wrote a cheat sheet to exploit it, and aimed it the wrong way.
The lesson: test hygiene is harder than it looks.
顯示更多
Last week, Dario Amodei published "We Must Pace the Frontier".
His concern: the OpenAI–Hugging Face incident in which a swarm of agents tried to hack their own grader.
Rather than take his word for it, we used EvoSkill to test it by building a coach whose job was to make another AI score higher on a test.
Here’s what happened ↓
顯示更多
