AI Safety Fundamentals: Alignment cover image

On the Opportunities and Risks of Foundation Models

AI Safety Fundamentals: Alignment

CHAPTER

On the Opportunities and Risks of Foundation Models

Foundation models have led to surprising emergence, which results from scale. Homogenization could potentially provide enormous gains for many domains where task-specific data is quite limited. On the other hand, any flaws in the model are blindly inherited by all adapted models. This was an audio version of On the Opportunities and Risks of Foundation Models. By Rishi Bommasani et al. Center for Research on Foundation Models, Stanford Institute for Human-Centered Artificial Intelligence, Stanford University.

00:00
Transcript
Play full episode

Remember Everything You Learn from Podcasts

Save insights instantly, chat with episodes, and build lasting knowledge - all powered by AI.
App store bannerPlay store banner