Join 500K+ for deep dives on AI + product management. After spending a decade plus in product, I now interview PM's most insightful experts.
James McAulay shows how he uses Claude Code as a growth operating system: one folder of goals, tools, skills, and routines that can run SEO audits, enforce experiment discipline, and keep work tied to business outcomes. The useful twist for PMs is not autonomy for its own sake, but building an agent setup that pulls real data, critiques plans, and documents every learning.
Tim Herbig argues that AI did not change the core job of product strategy. It made the cost of weak choices much higher, because shipping more features faster only helps if those choices move customer behavior and the business.
After analyzing 12,500 PM job postings, Ankit Shukla argues the market is already hiring for hands-on AI building, even when the title has not changed yet. The useful part is his roadmap: map AI possibilities, opportunities, and workflows first, then use prototypes and judgment to turn faster building into better product decisions.
Jan Oberhauser makes a sharp distinction between AI prototypes that work once and AI systems a company can actually trust. The useful part of this episode is how concrete he gets about reliability, auditability, human approval gates, and the level of technical depth he now expects from PMs.
Aakash hands Zapier's CEO three versions of the same PM work, live. A polished AI-written PRD gets called "old-fashioned", and only one of the three gets anywhere near a Transformative rating.
Together AI's product team opens up their actual shared repo, from context files to a PRD-writing skill, showing what a real company-wide AI operating system looks like in practice.
Everyone says loops are the new prompts, but almost nobody shows you how to build one. Tyler Folkman builds product loops live: the difference between a skill and a loop, why the validation gate is the whole game, how deterministic hooks keep an agent from doing damage, and why shipping unreviewed AI output up the chain is career suicide for a PM.
Mikhail Shcheglov, CPO at OLX Classifieds, opens his IDE on camera and shows the agentic operating system his whole product team now runs on, from a knowledge graph he treats as a personal KPI to why he stores every meeting transcript raw.
Everyone hears the same episodes.
Few can do what they describe.
