Artificial Intelligence and You 189 - Guest: Peter Norvig, AI professor/author/researcher, part 2
15 snips
Jan 29, 2024 In this intriguing discussion, Peter Norvig, a leading AI researcher and co-author of the seminal text on artificial intelligence, explores the evolving narrative around AI as its presence grows in society. He shares his thoughts on the movement to pause AI advancements and delves into the challenges of general intelligence within large language models. The conversation touches on the ethical complexities of AI in an accessible world, the dual nature of data, and wraps up with humorous insights on AI interpreting chicken sounds.
AI Snips
Chapters
Books
Transcript
Episode notes
Leaving Academia To Build At Scale
- Peter Norvig left academia because he wanted to build things at scale and teach broadly rather than chase tenure.
- He kept ties to universities, teaching online classes and returning periodically to academia.
Data Is Both Asset And Liability
- Norvig stresses that large datasets and human interaction signals powered major practical advances at Google.
- He later realized data can be a liability due to privacy risks and public trust issues.
Define Objectives Before Building Models
- Focus on defining the objective function and values when applying ML systems, not just algorithms or data.
- Decide what fairness and equity mean for your use case before building models.
