Experimentation, Causal Inference, and AI: Sean Taylor of OpenAI and VC Jesse Robbins at Data Council 2025

Heavybit by Jesse Robbins and Sean Taylor · · Article

"AI provides an opportunity to radically improve how we do things."

— Jesse Robbins

Experimentation, causal inference, and why AI generates more questions needing empirical answers — Jesse Robbins interviews Sean Taylor of OpenAI ahead of Data Council 2025.

In this Heavybit Library interview, Jesse Robbins sits down with Sean Taylor of OpenAI to preview the Data Science and Algorithms track for Data Council 2025. Their conversation focuses on experimentation, causal inference, and practical frameworks data teams can use to drive better product and business decisions — areas where data science creates measurable company value but often goes underappreciated.

The interview highlights featured speakers including Hadley Wickham on generative AI in data science workflows, Timothy Chan of Statsig on experimentation at scale, Joe Powers of Intuit on Bayesian A/B testing, and Bryan Bischof of Hex on ML engineering. Robbins frames the conference as a rare opportunity for isolated data science practitioners to validate their work by “peeking over each other’s shoulders.”