Heavybit Podcast Network Master Feed

On episode 46 of Open Source Ready, Brian Douglas and John McBride sit down with JT Archie. When an agent skips a command or grabs the wrong branch, a better prompt may only get you so far. JT explains how Steps uses structured pipelines to give agents clearer inputs and boundaries, then joins the hosts to explore SQLite, tool permissions, and the limits of fully automated development.

On episode 61 of Generationship, Rachel Chalmers speaks with Gabriela Moreira, co-founder and CEO of Quint, about building trust in AI-generated code. Gabriela explains how executable specifications help engineers explore software behavior and why passing tests can offer false reassurance. They discuss making formal methods more accessible while preserving the judgment and curiosity that good engineering requires.

On episode 18 of How It's Tested, Eden Ful Goh speaks with Jake Schuenke and Kevin Clark of Phantom about scaling quality assurance in the age of AI. They explain why faster code generation demands better testing, review, and release processes. From AI-assisted triage to real-device testing, they share how their team balances speed with the trust users place in a crypto wallet.

On episode 15 of Third Loop, the Progressive Delivery team discusses why AI makes human judgment more important than ever. Celebrating their book’s birthday, they revisit alignment, autonomy, abundance, and automation through the lens of AI agents. From evaluating models to recognizing good design, they explore what faster software creation still demands of people.

On episode 14 of Third Loop, the Progressive Delivery team speaks with Dr. Cat Hicks about what psychology says about building software with AI. Cat argues that the deeper you understand how humans learn and collaborate, the better you understand what makes AI work. Along the way: Descartes’ demon, Shrek the coding agent, and why “why do you need to do more?” is the question every engineering org should answer.

On episode 14 of Data Renegades, CL Kao speaks with Hannes Mühleisen, co-creator of DuckDB, about building databases that people actually want to use. Hannes explains how conversations with frustrated statisticians shaped DuckDB’s architecture and why useful software sometimes requires abandoning your own assumptions. They explore DuckLake, the Quack client-server protocol, and how a focus on human usability is paying off as AI agents begin choosing their own tools.

On episode 6 of Lab Notes, Amir Zohrenejad speaks with James Evans about a paradox of AI-powered science: faster answers may come with a narrower range of questions. They explore how diverse perspectives, productive disagreement, and new experimental capabilities could help AI expand scientific discovery. Evans also considers how research incentives shape what humans and machines choose to investigate.

On episode 45 of Open Source Ready, Brian Douglas and John McBride speak with Dotta about Paperclip, an open-source platform for managing teams of AI agents. They explore what happens when agents produce more work than humans can review, why permissions need to extend beyond a separate computer, and how engineering judgment changes when writing code becomes easier.

On episode 13 of Third Loop, the Progressive Delivery team speaks with Jennifer Riggins about why faster code doesn’t automatically mean better software. They explore how AI exposes organizational bottlenecks, why psychological safety matters, and what Spotify’s growing feature set reveals about user value. Jennifer makes the case for smaller experiments, stronger feedback loops, and keeping people at the center of delivery.

On episode 60 of Generationship, Rachel Chalmers speaks with Vasanth Mohan of SambaNova about when faster AI agents actually mean better work. They explore why token consumption and code volume are poor substitutes for business value, how inference latency interrupts productive workflows, and why engineering taste still matters. Sometimes the most valuable thing an engineer can produce is less code.