Time: 6:30-8:00 p.m., Wednesday, October 28th

Join us on: Zoom (password: Seminar) or YouTube Live

Speakers: JohnMark Taylor

Graphics: Olivia Foster Rhoades

You may not have heard of “artificial neural networks”, but if Facebook has ever automatically tagged you in a picture, you’ve definitely used one. Many experts once thought certain challenges in artificial intelligence wouldn’t be solved for decades: writing humanlike prose; recognizing people, places, and things; beating top human players at games like Go and Starcraft. In a few short years, however, all of these problems (and more!) have arguably been solved by AI algorithms that–believe it or not–are all built from the same simple building blocks. In this talk, you’ll learn what these building blocks are, how they’re built into systems that beat experts at some tasks, where they’re nonetheless beaten by toddlers at other tasks, and why society needs to pay attention.