What's The coolest topics In theoretical ai you've Seen that aren't completely unrealistic in a ai system?

I’ll add to John’s post, because I too think the work is facinating. Here are the latest evolution, updates, results etc of the Google Project: Accelerating Mathematical and Scientific Discovery with Gemini Deep Think Towards Autonomous Mathematics Research - The ArXiV Paper Of course there is the Erdo’s problems that the system solved: Semi-Autonomous Mathematics Discovery with Gemini: A Case Study on the Erdős Problems and Feng’s Eigenweights: Towards Autonomous Mathematics Research