Presenter
Yea-Hung Chen, UCSF Library
Session Overview
In this session, the third in the Library’s AI
in Your Toolkit series, we will discuss the extent to which AI tools
provide accurate information or offer sound suggestions. We will begin
by presenting notable examples of hallucinations—false or misleading
information—and summarizing research on the topic. We will offer some
possible techniques for minimizing hallucination. Finally, we will open
the floor for you to offer reactions to several guided questions.
Learning Objectives
- Summarize some of the research on hallucination
- Apply prompt-engineering techniques to minimize hallucination
- Formulate strategies for assessing AI output