Bay Area Tech Economics Seminar with Tom Cunningham, METR

Speaker: Tom Cunningham, Data Scientist, METR (previously at OpenAI)

Abstract: We discuss stylized facts about the capabilities of recent AI models and how they compare to human capabilities. We particularly draw attention to a common latent factor of capabilities among AI models, and the fact that this latent factor is different from the distribution of human capabilities. We discuss various candidates for the “missing capability” that would unlock a greater economic impact.

Speaker Bio: Tom has worked as a senior data scientist at Meta, Twitter, and OpenAI, on a range of topics, including strategy, pricing, anti-trust, advertiser behavior, and auction design, network effects (experimentation, modeling, tipping points), many aspects of experimentation, LLM model behavior, and content moderation.

Logistics details for each session will be provided to registered participants.

Sign up here to be included on the mailing list and learn about future events.

For additional information and abstracts from past talks, please click here.

Note: The BATES events are completely free to attend. If you’d like to support future events and our research, please consider making a gift to Stanford Data Science: https://bit.ly/supportSDS.