I'll describe the principles underlying the development of Aristotle, an AI theorem prover trained via reinforcement learning on Lean proofs, that acheived a gold medal equivalent score on this years IMO problems, has been used to solve open problems and autoformalize new papers in real time. I'll give a demo of how this can be accessed and show some examples of projects making heavy use of this technology, in particular those using it for novel mathematical proofs and for library development. I'll also discuss where (formal) mathematics might be headed as these tools continue to develop and evolve, and how working on mathematics might change as mathematicians incorporate these tools into their work process.
=== Hybrid talk ===
Join Zoom Meeting https://cam-ac-uk.zoom.us/j/89856091954?pwd=Bba77QB2KuTideTlH6PjAmbXLO8H...
Meeting ID: 898 5609 1954 Passcode: ITPtalk