Random thoughts, quick updates, and micro-posts
Breaking into product used to mean learning frameworks, roadmaps, and stakeholder management. In the AI era, that playbook is outdated. New PMs win by shipping, not studying. Build small AI products, run real experiments, and show you can create learning loops — not just slides. Your edge isn’t “PM knowledge.” It’s proof you can turn ambiguity into working systems. In 2026, portfolios beat resumes.
AI-first product management quietly kills the roadmap. When systems can discover opportunities, run experiments, and reprioritize in real time, a fixed 6-month plan becomes fiction. Most roadmaps today are comfort theater — a story for stakeholders, not a reflection of reality. The shift: from **committing to outputs** → to **committing to learning speed**. In AI products, the best roadmap is a live system — not a slide.
User research is becoming overrated in the AI era — but killing it is a mistake. With AI, PMs can rapidly prototype and run thousands of UserTesting tests to observe real behavior at scale. Behavior > opinions. But data shows *what*, not always *why*. Sometimes one deep conversation reveals the hidden variable behind millions of data points. AI finds patterns. Humans find meaning. User research isn’t gone — its role shifted to explain the data.
This is my first thought! 🎉 Testing out the new micro-blogging feature. I can now post quick updates without needing to git push from my computer. Pretty cool! Check out my site: https://thucldnguyen.com