Is there only one FairBot?
ai
A discussion on LessWrong is exploring whether there is only one version of the FairBot, a strategic agent used in a prisoner's dilemma tournament hosted by the Machine Intelligence Research Institute. The author notes that different mathematical definitions of the bot, including those from Vladimir Slepnev and James Payor, appear to be logically equivalent. This has led to questions about whether any unexploitable agent that cooperates with itself is effectively the same FairBot.
Source: https://www.lesswrong.com/posts/auAq7Rcstop3FBEob/is-ther...
Listen to this story
Hear this and more stories in a personalized audio briefing.
Open The Chonkerton