On product management, Stoic detachment, and what AI actually exposes
There is a version of a product manager that has always existed inside organizations, quietly doing damage. They show up to standups. They groom backlogs. They write tickets, run ceremonies, and facilitate alignment meetings that could have been emails. They protect their roadmaps like sacred documents and nod at whatever the CEO says in the all-hands. They are, in the most precise sense of the word, administrators.
For a long time, organizations tolerated this. The work got shipped. The process got followed. Nobody asked too hard whether the thinking behind it was any good.
AI is ending that grace period.
What product management actually is
The most undervalued thing a product manager does — the thing that separates the real ones from the drones cosplaying in the role — is decision-making. Not facilitation. Not documentation. Not stakeholder management. Decisions. And decisions require something that cannot be prompted into existence: an educated opinion, backed by evidence, with the courage to defend it and the humility to abandon it.
That last part matters. The courage and the humility have to coexist.
I've watched product organizations collapse inward because the PM layer became a translation layer — converting executive whims into engineering tickets, mistaking responsiveness for value, confusing velocity with direction. The PM who never pushes back isn't protecting anyone. They're accelerating in the wrong direction with better documentation.
The antidote isn't pushback, exactly. Pushback is reactive. What the role actually demands is something more fundamental: question everything. Including, and especially, yourself.
Early in a career, political capital is scarce. You can't walk into your first month and veto the VP's pet project. But you can ask why. You can ask have we considered. You can ask what problem does this actually solve. If you're not doing that in the first week, you're not doing the job — you're just attending it.
The roadmap is not the asset
I've killed entire roadmaps. Not trimmed them — killed them. Looked at months of planning and concluded that the right move was to start over or dramatically compress what we thought we were building.
It's uncomfortable the first time. Then it becomes the job.
What I learned from doing it is that the value was never in the roadmap. The roadmap is an artifact. It's a document that describes what you currently believe to be true about the problem, the user, and the business. When those beliefs change — and they always change — the document should change with them. Clinging to it isn't commitment. It's ego.
The ancient Stoics had a framework for this that I find more useful than most product methodology: they drew a hard line between what is inside your control and what is not. The outcome of a feature launch is not inside your control. Market timing, user behavior, competitive response — none of it is yours. What is yours is the quality of the thinking that led to the decision. The rigor of the question you asked. The honesty of the answer you gave.
A product manager who conflates those two things — who ties their professional identity to outcomes rather than judgment — becomes fragile. They defend bad ideas past the evidence. They can't kill their own work. They mistake the map for the territory and refuse to update when the territory changes.
The ones who separate them become durable. They build capital not by being right every time, but by being known as someone who thinks. Someone who has a point of view, can back it with real ideas and real evidence, and can change that point of view when the evidence demands it without treating the change as defeat.
That is the asset. Not the roadmap. Not the shipped feature. The demonstrated quality of judgment over time.
What AI actually exposes
Here is the thing about AI tools that most product discourse gets wrong: the threat is not that AI replaces product managers. The threat is that AI reveals which product managers were replaceable all along.
Ticket writers are in trouble. Coordinators and facilitators whose primary value was in organizing information and running process — their roles will shrink or disappear. AI does that work faster, cheaper, and without ego. If that description makes you nervous, it's worth sitting with why.
Where AI cannot go — not now, not in any near-term future I can see — is into genuine instinct. Not pattern-matching on training data, which AI does extraordinarily well. I mean the calibrated intuition that comes from spending years inside a specific domain, feeling where things are broken before the data confirms it, knowing why a feature that looks right on a spec sheet will feel wrong to the user at 11pm on a Sunday.
That kind of instinct is not transferable. It's not documented. It lives in the gap between what can be written down and what can be known. It's why the best product managers I've observed tend to stay in their space — not because they lack ambition, but because depth compounds in ways that breadth does not.
The future of this role, as I see it, belongs to builders with domain expertise. People who can open a tool, spin up a prototype, test an idea before it becomes a ticket — and who have spent enough years in a specific vertical that their instincts are actually calibrated. Sports betting. Consumer banking. Trading apps. Health platforms. The niches where the users are specific, the stakes are real, and the generic solution is always wrong.
Those people become more valuable as AI gets better, not less. Because AI needs direction. It needs someone who knows which questions are worth asking.
The question is the job
This is what I keep coming back to: AI is extraordinarily good at answering questions. It will only get better at that. What it cannot do is know which questions matter.
That is the job. It always was. We just had enough ambient complexity in the process — the meetings, the tickets, the ceremonies — to obscure it.
The product managers who thrive in the next decade will be the ones who internalized that early. Who built their identity around the quality of their inquiry rather than the ownership of their roadmap. Who can sit across from a CEO, push back with evidence and humility, kill their own ideas when the data demands it, and walk away with their judgment intact regardless of how the outcome lands.
Not because they read it in a framework.
Because they practiced it until it became instinct.
Rainer Floeter is a Staff PM at theScore BET. He writes about sportsbook product, the business of sports betting, and what it means to build well.