The Test Set: Now on YouTube + a look at what’s next

2025-12-04

Four people sitting around a table with a title overlay, "the test set."

If you’re the kind of person who gets amped about a perfectly tidy dataset or a beautifully written function, we have great news: The Test Set, Posit’s video podcast for data science junkies and anomaly hunters, is officially on YouTube.

WATCH ON YOUTUBE

We’re still early in the journey — 11 episodes in — and it’s already been a blast. And now we’re bringing you more insightful and entertaining conversations with the people behind your favorite tools, plus what’s what in the world of statistical analysis, scientific computing, and open-source data science.

Here’s a sneak peek of what’s to come in 2026 with a stacked lineup of guests who are shaping the future of the industry:

Tristan Handy — CEO & Founder, dbt Labs
Nellwyn Thomas — VP of Data, Shopify
Benn Stancil — Founder of Mode
Charlie Marsh — Founder & CEO, Astral
Soumith Chintala — Former VP & AI Fixer at Meta, now investor
And many more we can’t wait to share

Missed the first episodes? Here’s the TL;DR.

If you’re new to the podcast (or just need a recap), The Test Set is loaded with behind-the-scenes perspectives spanning everything from the inadvertent blossoming of data science to the future of open source. Hosts Michael Chow, Hadley Wickham, and Wes McKinney sit down with industry players and pioneers who’ve helped build modern data science as we know it and are imagining what comes next. And while we’ve been celebrating talented developers and educators at Posit, we are thrilled to highlight folks from across companies and industries in the coming months!

Here are some highlights from previous episodes:

- Ever wonder how the tidyverse came to be? Hadley Wickham takes us on a journey from his early days creating databases to the inspirations behind his groundbreaking work in R. (Link)
- The creator of pandas, Wes McKinney, shares his frustrations that fueled the creation of one of Python’s most beloved data libraries and sheds light on the hidden side of open source: the hustle. We also learned about how his passion for video game speedrunning influenced his approach to building open-source tools. (Part 1 Link, Part 2 Link)
- How do you teach data science when AI can write code? Mine Çetinkaya-Rundel shares her “whole game” approach and experiments using LLMs to provide instant student feedback. (Link)
- Roger Peng shares his journey as an early R adopter and online educator. He explores the evolution of data science from its origins and why teaching remains at the heart of his work. (Link)
- Michael Chow shares his journey from pursuing a PhD in cognitive psychology to becoming a Principal Software Engineer at Posit. We explore the “behind-the-scenes messiness” of building data tools, and how his passion for making data easier and more fun led to open-source projects like Great Tables. (Link)
- Julia Silge, renowned astronomer-turned-data science leader, offers deep insights into the unique challenges of data science workflows, the development of Positron, and the impact of the LLM era. (Part 1 Link, Part 2 Link)
- James Blair traces his “delightfully non-linear path” unique career path into the data science world and offers a fascinating look at the future of AI-powered tooling within the Posit ecosystem. (Part 1 Link, Part 2 Link)
- Kelly Bodwin discusses her path from English to Associate Professor of Statistics and Data Science at Cal Poly, and what it takes to keep a programming community weird in the best possible way. (Link)

Subscribe, watch, listen, engage

The Test Set was created for people who get a dopamine rush from a clever plot, shaving minutes off a pipeline run, or when a model converges on the first try. If you identify with this, you’ll feel right at home on The Test Set. Catch the latest episodes, and make it part of your routine — now streaming on YouTube and wherever you get your podcasts.