Product Management at AI native companies
Let me tell you about the most fascinating identity crisis happening in tech right now: Product Management is evolving so fast that by the time you update your LinkedIn title, it’s already obsolete. And by “evolving,” I mean the role is basically consuming every other role like some kind of corporate Pac-Man while simultaneously questioning its own existence.
Here’s what’s happening to PMs in the age of AI, and it’s both terrifying and exhilarating in the way that watching your career transform in real-time tends to be.
The Old World (RIP)
Remember when we had this perfectly choreographed corporate ballet? The PM would write requirements. The designer would design things. The engineer would engineer things. Everyone stayed in their lane like well-behaved highway drivers, and occasionally they’d have “cross-functional meetings” where they’d argue about whose turn it was to use the conference room.
It was inefficient, sure, but at least everyone knew their place in the organizational chart. You could draw boxes around people. Managers love drawing boxes around people.
Enter the Talent Stack Implosion
Now? Now we’re watching what I call the Great Talent Stack Collapse of 2025, and it’s making traditional org charts look like abstract art.
PMs are building prototypes. Not wireframes. Not “clickable mockups.” Actual working prototypes with AI assistance that would’ve taken a team of three engineers six weeks to build in 2019.
Engineers are deciding requirements. They’re talking to customers. They’re doing the thing we used to call “product strategy” except nobody told them they needed an MBA first.
Designers are shipping code to production. Real code. The kind that used to require a deployment pipeline, three approval gates, and a ritual sacrifice to the git merge gods.
What’s Really Happening Here
The secret that nobody wants to admit out loud: all those carefully constructed role boundaries were mostly about managing scarcity. Scarcity of tools, scarcity of knowledge, scarcity of time.
When it took six months to learn enough JavaScript to be dangerous, you needed specialists. When design tools required a degree in graphic design to operate, you needed specialists. When understanding customer needs required extensive market research budgets, you needed specialists.
AI didn’t just make these tools easier. It made the whole concept of specialization start looking suspiciously like an HR convenience rather than a business necessity.
The Multidimensional Builder Emerges
So what replaces the PM/Eng/Designer trinity? I’m betting on something I’ll call the Multidimensional Builder, because I need something to put in my LinkedIn headline.
These are people who can:
- Prototype an idea Tuesday morning
- Validate it with actual code Tuesday afternoon
- Talk to three customers Tuesday evening
- Ship a beta Wednesday
- Decide it was a terrible idea Thursday
- Start over with something better Friday
They’re not “good at everything.” They’re good enough at enough things that the collaboration tax—that invisible overhead from coordinating between specialists—basically drops to zero.
The Uncomfortable Truth
Here’s where I’ll make enemies: those incredibly specific job titles we’ve been collecting like Pokemon cards? “Senior Principal Staff Product Manager III” or “Lead Strategic Design Engineer Architect”? They’re starting to look like elaborate participation trophies from the Before Times.
Not because the skills aren’t valuable. But because the boundaries between the skills are evaporating faster than your manager’s patience during a failed sprint demo.
The Part Where I Tell You What to Do
If you’re a PM who can’t build, learn to build. If you’re an engineer who won’t talk to customers, start talking to customers. If you’re a designer who thinks code is beneath you, that’s fine, but the designer who doesn’t think that just shipped your feature while you were picking fonts.
The good news? AI makes learning this stuff approximately 47 times easier than it used to be. (That’s not a real statistic. I made it up. But it feels right, doesn’t it?)
The better news? The people who embrace this aren’t just more employable. They’re more effective. They can think of an idea and bring it to life without playing telephone through three departments.
The Future Is Weird and Fast
Five years from now, I predict job postings that say “Builder” and list literally every skill from 2024’s entire product org. And the crazy part? People will actually have those skills.
Not because humans got smarter, but because AI became the ultimate force multiplier for the jack-of-all-trades. It turned “kinda good at everything” from a liability into a superpower.
The specialists will still exist, of course. But they’ll be working on problems so gnarly that even AI throws up its hands and says “you know what, you handle this one.”
For everything else? There’s going to be some Builder in their pajamas on a Tuesday morning, shipping something that used to require a team of twelve.
Welcome to the collapse of the talent stack.
It’s going to be chaos.
I can’t wait.
One More Thing
If you want to dive deeper into how product management is evolving, check out this video by Claire Vo, CPO at LaunchDarkly. She articulates this transformation way better than I do, and with significantly fewer made-up statistics.
P.S. - If you’re a manager reading this and panicking about how to organize your teams, congratulations! You’ve identified the next blog post topic. The answer rhymes with “stop organizing them so much.”