The best data science podcasts in 2026 are Data Skeptic, Super Data Science, Practical AI, The TWIML AI Podcast, Gradient Dissent, Eye On A.I., Machine Learning Street Talk and Latent Space. This list is about data science practice: statistics, applied machine learning, data careers and putting models into production. If you want ML research and model-building, our best machine learning podcasts list is the better fit.
Built This Week is our show. It isn't a data science podcast, so we list it at the end as "Also from us," unranked. Here's how we picked the rest.
Kyle Polich hosts Data Skeptic, which features interviews and discussion on data science, statistics, machine learning and AI, all from the perspective of critical thinking and the scientific method. It publishes every two weeks or so, as of October 2026, with episodes of about 35 minutes, and its feed goes back to May 2014. Recent episodes run in themed series, such as recommender systems and their optimization goals.
It is the most "data science" show on this list: good for analysts and data scientists who want to question claims and methods, not just hear about new models.
Start with: N-Beats (3.8K views on YouTube as of October 2026). Full profile of Data Skeptic.
Dr. Jon Krohn hosts Super Data Science, covering machine learning, AI and data career topics from academia and industry. It publishes two to three times a week, as of October 2026, with episodes of about 48 minutes, and its feed goes back to September 2016, with more than 1,000 episodes. The schedule mixes guest interviews with shorter solo episodes and monthly "In Case You Missed It" recaps.
It suits data scientists who want a steady stream of industry interviews and career context.
Start with: 953: Beyond "Agent Washing": AI Systems That Actually Deliver ROI (207K views on YouTube as of October 2026). Full profile of Super Data Science.
Daniel Whitenack and Chris Benson have hosted Practical AI since July 2018. The show's focus is productive implementation of AI, from machine learning and MLOps to LLMs, through discussion and guest interviews. It publishes weekly, as of October 2026, and episodes usually run about 47 minutes; it is rated 4.3 out of 5 from 199 ratings on Apple Podcasts, as of October 2026.
It is a good fit for practitioners who ship models and want grounded, implementation-first conversation.
Start with: Hermes Agent: Agents that grow with you (2.4K views on YouTube as of October 2026). Full profile of Practical AI.
Sam Charrington has hosted The TWIML AI Podcast (formerly This Week in Machine Learning & AI) since May 2016. It interviews ML and AI researchers and practitioners for an audience the show describes as researchers, data scientists, engineers and tech-savvy business and IT leaders. It publishes every two weeks or so, as of October 2026, with episodes of about 59 minutes; it is rated 4.7 out of 5 from 426 ratings on Apple Podcasts, as of October 2026. A recent episode on ranking models optimizing the wrong thing is typical of its applied side.
Start with: Are Large Language Models a Path to AGI? with Ben Goertzel - 625 (42K views on YouTube as of October 2026). Full profile of The TWIML AI Podcast.
Lukas Biewald, founder of Weights & Biases, hosts Gradient Dissent, launched in March 2020 and produced by Weights & Biases. Conversations focus on AI in industry and the details of bringing models into production. It publishes about monthly, as of October 2026, with episodes of about 58 minutes; the most recent episode in its feed is from August 2026. It is rated 4.8 out of 5 from 70 ratings on Apple Podcasts, as of October 2026.
Start with: Exploring PyTorch and Open-Source Communities with Soumith Chintala. Full profile of Gradient Dissent.
Craig S. Smith, a longtime New York Times correspondent, hosts Eye On A.I., talking to people making a difference in AI and putting incremental advances into a broader context. It publishes two to three times a week, as of October 2026, with episodes of about 55 minutes, and its feed goes back to October 2018. Recent titles cover enterprise uses, such as AI agents for procurement, and a test of 65+ models for cost.
Start with: The Mastermind Behind GPT-4 and the Future of AI | Ilya Sutskever (405K views on YouTube as of October 2026). Full profile of Eye On A.I..
Machine Learning Street Talk hosts in-depth discussions with leading figures in AI, covering current affairs in AI, cognitive science, neuroscience and philosophy of mind. It publishes weekly, as of October 2026, with long episodes of about 79 minutes, and its feed goes back to April 2020. It is the most theory-heavy pick here, for data scientists who want to understand why methods work.
Start with: The Ghost in the Machine — Noam Chomsky (952K views on YouTube as of October 2026). Full profile of Machine Learning Street Talk.
Latent Space is led by swyx with co-hosts Alessio and Vibhu, and has been running since February 2023. It covers AI engineering, foundation models, agents and infrastructure through builder interviews plus news and paper discussion. It publishes two to three times a week, as of October 2026, and episodes usually run about 85 minutes. Recent guests include the CEO of Modal and an SF Compute founder, so it is useful for data people moving toward AI and data infrastructure.
Start with: Claude Code's Next Era — Thariq Shihipar, Anthropic. Full profile of Latent Space.
Built This Week, hosted by Jordan Metzner and Sam Nadler, is a weekly AI show rather than a data science one; the closest episode is Episode 27 with Ben Lerner of Espresso AI, where Sam built a data warehouse waste quiz.
Schedules and lengths are as of October 2026. Related lists: the best machine learning podcasts, the best AI podcasts for developers and the best coding podcasts, or see all our podcast lists.
Data Skeptic, Super Data Science, Practical AI, The TWIML AI Podcast, Gradient Dissent, Eye On A.I., Machine Learning Street Talk and Latent Space. The first three are the most focused on day-to-day data science and applied ML practice.
For applied ML, Super Data Science (Jon Krohn), Practical AI (Daniel Whitenack and Chris Benson) and The TWIML AI Podcast (Sam Charrington) are good starting points. For research-heavy ML, see our separate list of the best machine learning podcasts.
Data Skeptic, hosted by Kyle Polich, covers data science, statistics and machine learning from the perspective of critical thinking and the scientific method. It publishes every two weeks or so, with episodes of about 35 minutes, as of October 2026.
Super Data Science with Dr. Jon Krohn covers machine learning, AI and data career topics from academia and industry, publishing two to three times a week as of October 2026.
