I Spent 15 Years as a Product Manager. The Job I Trained For Is Gone.

I got my first product job fifteen years ago, in a large European bank. My job, as it was explained to me, was to write documents that described what engineers should build, and then attend meetings about those documents. I got good at it. Embarrassingly good. I could produce a forty-page functional spec that nobody would ever read past page six, and I knew nobody would read it, and I wrote page seven anyway.

That job is disappearing. Not slowly, not "the role is evolving" — disappearing. And I want to explain why I think that's the best thing that has happened to this profession since someone decided PMs shouldn't report to marketing.

The uncomfortable math

Here's what my weeks actually looked like for most of my career, whether in banking, insurance, or later in fintech: maybe sixty percent of my time went into producing and synchronizing documents. PRDs, status updates, meeting notes, steering committee decks, release notes. Another chunk went into being a human API between teams. What was left — the part where I actually thought about users and made product decisions — was maybe a day a week. On a good week.

Nobody designed it that way. It's just what the role sedimented into, because documentation and coordination were genuinely hard, and someone had to do them.

Then, over the last couple of years, that sixty percent collapsed. An AI drafts a solid PRD in four minutes from my messy notes. It writes the stakeholder update from my roadmap board. It synthesizes twelve interviews before my coffee is done. The work that filled my calendar for a decade became nearly free.

You'd think PMs would be celebrating. Some are. But I've watched a lot of them quietly panic instead, because when the documentation work evaporates, a harder question surfaces: what exactly do you do now?

What's replacing us

The answer, I think, has a name: the product builder.

The distinction isn't cosmetic. A product manager, in the classic sense, coordinates other people who build. A product builder takes an idea from zero to one with minimal dependencies — talks to users on Monday, prototypes on Tuesday with tools like Lovable or Claude, tests on five humans by Thursday, and walks into Friday's meeting with a working thing instead of a deck about a thing.

I've seen what happens in a room when someone does this. A prototype that works beats a presentation that impresses, every single time. The person with the clickable thing sets the agenda. Everyone else is reacting to it.

The bottleneck moved. It's not engineering anymore — it's deciding what deserves to exist.

The part nobody tells you

Here's what I got wrong at first, and what I see many PMs getting wrong now: they think the transition is about learning the tools. It isn't. The tools take a weekend. I've watched people master every AI tool on the market and still ship garbage — beautifully formatted, confidently worded garbage.

Because the AI has a flaw that took me a while to name: it never tells you your idea is bad. It takes whatever you give it and makes it look finished. If your problem statement is fuzzy, you get a polished document about a fuzzy problem. The draft was never the hard part. The hard part was always the judgment — knowing which assumption is load-bearing, which metric is vanity, which stakeholder request to kill. That part didn't get automated. It got more valuable, because it's now the only part that's scarce.

So the builders who win aren't the ones who delegate their thinking to the machine. They're the ones who built a workflow around a simple division of labor: the AI drafts, they attack the draft, and then they decide. Fifteen years of writing specs taught me less about product than two years of tearing apart AI drafts has.

What I did about it

At some point I stopped waiting for someone to hand me the operating manual for this new job, and I built it myself — in Notion, because that's where my work lives. A system where ideas get scored, evidence gets collected, specs get drafted by AI and then attacked by AI, prototypes get built in afternoons, and one weekly ritual holds the whole thing together. Nothing advances without evidence or a decision.

I built it for myself. Then I realized the transition I went through is the one an entire profession is going through right now, mostly without a map. So I cleaned it up, filled it with a worked example from start to finish, and made it something you can duplicate in thirty seconds.

It's called the Product Builder OS. If you're a PM feeling the ground shift — or someone who never had the PM title but ships product anyway — it's the system I wish someone had handed me two years ago.

The job I trained for is gone. The one that replaced it is better. But only if you make the jump deliberately, with your judgment intact — because that's the one thing the machine still can't draft.