MoMath: The Mathematics of AI

The National Museum of Mathematics
225 5th Ave, New York, NY 10010

Artificial intelligence is breaking new ground every week.  It writes essays, codes, drives cars, and identifies objects in photos.  Much of this is based on neural networks, which coarsely mimic the brain.

Take a closer look at the computations behind AI in this presentation led by Samuli Siltanen, Professor of Industrial Mathematics and Vice Dean of the Faculty of Science at the University of Helsinki, Finland.

Amazingly, we will see that neural networks use only elementary school math to work their magic, that orienteering strategies have implications on learning, and that the automatic “learning” that AI code does is akin to a mountaineer looking for the lowest valley in thick fog.