Mitigating the Risks and Harms of Large Language AI Models

Mitigating the Risks and Harms of Large Language Models, with Dr. Kush R. Varshney

Wednesday, October 25 @ 7:00 PM

REGISTER

While ChatGPT and other large language models have enthralled the world with their capabilities, using them in consequential applications is not without clear and present risks. We will talk about socio-technical harms that these technologies are enabling like misinformation, leakage of private information, unfairness, unwanted stereotyping and bullying. Then new safety mechanisms being developed to mitigate these risks will be discussed.

Kush R. Varshney, PhD, is a Distinguished Research Scientist and Senior Manager with IBM Research at the Thomas J. Watson Research Center, Yorktown Heights, NY. His book, Trustworthy Machine Learning, is available as a free PDF and html: http://www.trustworthymachinelearning.com  or at-cost paperback: https://www.amazon.com/dp/B09SL5GPCD

This lecture is part of the Library’s series “AI: Ready or Not “which is made possible by the generosity of the Friends of the Ridgefield Library.