Really liked the framing around Super ICs vs AI PM builders as two distinct paths. The shift from 'creating output' to 'evaluating AI generated output' is spot on and something I've noticed first-hand when using Claude for research syntheis. What gets overlooked tho is how this evaluation skill is way harder to teach than execution was. Back when I was hiring junior PMs last year, most bootcamp grads could follow a PRD template but almost none could critique one meaningfully, and now that's table stakes.
insightful for an aspiring pm
Really liked the framing around Super ICs vs AI PM builders as two distinct paths. The shift from 'creating output' to 'evaluating AI generated output' is spot on and something I've noticed first-hand when using Claude for research syntheis. What gets overlooked tho is how this evaluation skill is way harder to teach than execution was. Back when I was hiring junior PMs last year, most bootcamp grads could follow a PRD template but almost none could critique one meaningfully, and now that's table stakes.
The bar has gotten higher for junior PMs for sure. The good thing is LLMs and lots of free resources make it easier to learn these skills.