I lost an AI box experiment against PatrickRobotham with me as the AI today on irc. If anyone else wants to play against me then PM me here or contact me on #lesswrong.
Failing to convince your jailer to let you out is the highly likely outcome, so it is not very interesting. I would love to hear about any simulated AI winning against an informed opponent.
If it's worth saying, but not worth its own post (even in Discussion), then it goes here.