Vanishing Gradients podcast: agents are like coffee — they help you make stupid mistakes faster
hugobowne · x · 2026-10-05
On the latest Vanishing Gradients episode, Hugo Bowne-Anderson talks with scikit-learn creator Gaël Varoquaux (Probabl), Shipra Arora (Bain) and Luca Fiaschi (PyMC Labs) about agentic data science. Varoquaux's quip: agents are like coffee — they get you to do stupid things faster with more energy.
Key points:
- Experienced data scientists get more done with agents; people without expertise just think they do — code running and accuracy looking good doesn't mean the analysis supports the conclusion.
- Verification approaches: programmatic checks, agents critiquing agents, and where human judgment still matters.
- New possibilities when business users can question models and run simulations themselves, letting data teams finally tackle their backlogs.
Related event: scikit-learn Founder Discusses Agentic Data Science on Podcast(2 posts)→
More from coding & agent
- Viral meme: AI agents with full computer access will indeed modify your files — smtabatabaie · 2026-10-05
- Ethan Mollick: We need team email systems built for humans and AI agents to share — emollick · 2026-10-05
- 7 trade-offs every AI engineer makes, from accuracy-vs-speed to privacy-vs-personalization — goyalshaliniuk · 2026-10-05
- GPT-6.1 Sol + Blender MCP turns Gundam photos into a textured 3D model in under 10 minutes — sidahuj · 2026-10-05
- How this team shipped 2,500 PRs to production last month: poteto shares the full workflow — gaganghotra_ · 2026-10-05
- Addy Osmani: Equip your coding agent with tests so it can check its own work — addyosmani · 2026-10-05