Wow, Cognizant AI Lab seems to have perfectly solved Tower of Hanoi that is so hard for LLMs (as Apple's Illusion of Thinking paper showed) with a 'society' of agents approach.
Loading thread detail
Fetching the original tweets from X for a clean reading view.
Hang tight—this usually only takes a few seconds.