Last week, one line from Marty Cagan stayed with me:
"AI is having a very different impact… it ranges from your job is in serious trouble to you're making more money than you ever dreamed of."
Most people read that as a warning. I read it as an opening.
"Will AI take the PM's job?" is the wrong question.
The right one: how much more powerful does AI make a PM who already focuses on outcomes?
My answer: significantly. More than most people realize.
But first you have to understand what a PM week actually contains. There are two kinds of work inside the role. One is coordination — meeting notes, ticket management, sprint documentation, status reports, Jira updates. The other is the actual job — which problem are we solving, does the user really want this, is this prioritization right, where does the technical cost land?
AI is taking over the first category. Or it's about to.
For an outcome-driven PM, that isn't a loss. It isn't a gap. It's reclaimed time.
Let me make this concrete.
Think about a PM who spends several hours a week taking meeting notes, writing status updates, maintaining ticket hygiene. When AI takes those tasks, what does that PM gain? More time talking with users. Real space to work through a hard problem with the technical team. Room to question why a feature entered the backlog in the first place — not just whether it shipped.
When the coordination noise clears, the actual product thinking has space to breathe.
Reading that as "AI is taking my job" instead of "AI just cleared my schedule for the work that matters" is a fundamental misread of the tool.
There's a second thing AI gives PMs that gets less attention: option generation.
AI can now synthesize user research, draft specs, run competitive analysis, generate A/B test hypotheses — and do most of it quickly, at reasonable quality.
But "which of these is right?" is still the PM's job.
Does this feature actually create value for the user? Should we move this metric, or solve that other problem instead? Does this technical debt get paid now? AI can't answer those questions — because those decisions require context, strategy, and user understanding that lives in the PM's head.
As AI generates more options, the value of deciding between them goes up. The PM who makes those decisions well becomes more valuable, not less.
Marty Cagan has said for years that the real PM job is making engineers and designers better — clarifying the right problem, protecting their motivation, calling out the bad idea early, defending the good one at the right moment.
AI can't do that work.
But AI gives PMs the time to do it. As the coordination load lifts, the PM can have a deeper conversation with an engineer about why something is being built. Can run longer user sessions. Can sit with a decision long enough to actually think it through.
That's the PM Marty Cagan has always described. AI makes that PM more available — more present, more capable, less buried.
Here's what I keep seeing: the most valuable thing in a product team is the ability to make the right call at the right moment. To not solve the wrong problem. To see what the user actually wants, not what they said they wanted.
That capability doesn't come from AI. It can be developed, but it can't be generated.
When an outcome-driven PM starts using AI tools seriously, the effect compounds: less time on coordination, more on real product thinking. Faster analysis, sharper decisions. And the team around them notices — because that PM shows up with more mental bandwidth, more often.
That's the amplification.
Marty Cagan called it divergence. He's right.
But I read divergence not as a threat, but as clarification. AI removes the coordination layer, and the actual PM work becomes more visible. For the PM who's already focused on outcomes, this is the best environment there's ever been.
The tools are ready. The time is opening up.
One question left: what are you going to do with it?