AIIP Webinar: Using AI to Summarize & Synthesize Information

Using AI to Summarize & Synthesize Information

This event is open to members and non-members of AIIP.

Daniel Barulli will show us how to use AI to summarize and synthesize information across a variety of sources (public or private) to save time and/or extract relevant information. He will provide a brief overview of how the underlying AI works (primarily large language models (LLMs) and knowledge graphs). The webinar will describe some of the possible tiers of summarization and when each might be appropriate, including:

a) using generic models (such as GPT-4) to summarize information,
b) using specialized models to summarize information, and
c) fine-tuning your own models to summarize information.

Daniel will explain what "information" means in this context (text, images, audio and video media). He will also briefly touch on the tools available to do this, both proprietary and open source, and discuss the trade-offs involved for each, particularly data security/privacy considerations, and the risk of "hallucination." And finally he will provide demos of the different options working in different contexts, including some historically challenging ones, e.g. technical academic literature.

Meeting is on Wednesday, May 15, 2024, from noon-1pm U.S. Eastern time.

Register at https://us02web.zoom.us/webinar/register/WN_8wbLYP8hQ7ua8K9E53Ilow