Obviously everyone is super excited about LLMs right now, and while there's a large element of hype in the popularity they are also genuinely useful. In this talk I'll give a round up of data science things that I've found LLMs particularly useful for, broken up into three broad categories: writing code, writing prose, and rectangling fundamentally non-rectangular data (e.g. test, images, videos, audio).
Hadley Wickham is Chief Scientist at Posit, where he shapes the future of data science tools and workflows. Creator of the tidyverse and numerous influential R packages, he's transformed how people work with data in R. A statistics Ph.D. from Iowa State, he combines deep technical expertise with a passion for making data science more accessible. When not coding, he's teaching at Rice University or advancing data visualization techniques through his widely-cited research.