What Do We Want From LLMs?: Agents and Tool Use
This video explores why testing Large Language Models (LLMs) with simple factual questions misses the point of their true potential. Instead, it demonstrates ho
What Do We Want From LLMs?: Agents and Tool Use
It seems like every couple of weeks there’s a brand new large language model coming out. And when that happens, a lot of people jump online and start to test these models by asking, Vibe Questions such as, (screen whooshes) “Who Won the World Series,” and questions like that. The interesting thing about this, of course, is that that’s a terrible question to ask of a large language model, because large language models, they’re trained on general data, on data from the internet.
Chapters
- 00:00 - Introduction
- 01:32 - PartyRock Demo
- 03:03 - Building agents with Bedrock
- 12:33 - Outro
Links & Resources
This post is licensed under CC BY 4.0 by the author.