BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Math for AI Safety - Lionel Levine (Cornell University)
DTSTART:20241031T103000Z
DTEND:20241031T111500Z
UID:TALK220735@talks.cam.ac.uk
DESCRIPTION:AI holds great promise and\, many believe\, great peril. What 
 can mathematicians contribute to ensuring that promise is fulfilled\, and 
 peril avoided? Using Activated Random Walk as a running example\, I&rsquo\
 ;ll illustrate the capabilities and the shortcomings of large language mod
 els on reasoning and coding tasks. I'll highlight some research directions
  that might help to align future AI systems with human well being.
LOCATION:Seminar Room 1\, Newton Institute
END:VEVENT
END:VCALENDAR