The role of a Product Manager is evolving fast. With large language models becoming more capable every quarter, the skills that define great PMs are shifting underneath us.
What's changing
Traditional PM work—writing specs, triaging bugs, running standups—is increasingly augmentable by AI. The PMs who thrive will be the ones who lean into this rather than resist it.
Here are a few trends I'm watching:
- AI-assisted discovery: Using LLMs to synthesize user research, scan support tickets, and surface patterns humans miss.
- Prototype velocity: Tools like Cursor and Claude Code let PMs build functional prototypes without waiting for eng cycles.
- Strategy over execution: As execution gets cheaper, the premium shifts to taste, judgment, and knowing what to build.
What stays the same
Despite all the hype, some fundamentals aren't going anywhere:
- Customer empathy — No model can replace sitting across from a user and watching them struggle with your product.
- Cross-functional leadership — Aligning engineering, design, and business stakeholders still requires human nuance.
- Saying no — The hardest part of product management remains deciding what not to build.
My take
I think we're heading toward a world where every PM is a "technical PM" by default—not because they write production code, but because they use AI tools fluently enough to validate ideas independently. The bar for what one person can ship is about to go way up.