The PM job description is being rewritten. Most product managers haven't noticed.
Nikhyl Singhal, a product leader who has built teams at Meta and Google, put the forecast clearly this week: the next two years represent the most chaotic period in product management history. Approximately half of current PMs are at risk. Companies will shed around 30,000 PM positions and rehire roughly 8,000 new ones structured around AI-first ways of working.
The Restructuring Is Already Priced In
That's not elimination. That's replacement with a much smaller number of more capable people. The math is straightforward once you accept the premise.
AI can synthesize customer feedback, draft requirement documents, generate product specs, run competitive analysis, and track project timelines. The work that used to require a team of five PMs can be done by one PM with the right AI infrastructure and judgment to direct it.
The roles being shed are the coordination-heavy, requirements-documentation, stakeholder-communication versions of PM. The roles being created are closer to something between a product engineer and a business systems architect. Different inputs. Different outputs. Smaller headcount. Higher individual leverage.
The "fancy logo" resume matters significantly less than it used to. Demonstrated ability to work effectively with AI and willingness to fundamentally reimagine PM responsibilities matters more. The companies hiring now know this. Most of the PMs in the market haven't fully internalized it.
What Survives the Transition
Two things resist automation in product management: judgment and relationships.
Judgment about which problem is worth solving. Understanding of user psychology that comes from years of watching real people struggle with real products. The ability to be in a room with an executive team and read what's not being said.
Relationships with engineers, designers, and leadership that translate into operational trust. Trust is what gets a product shipped when the timeline is broken, priorities are conflicting, and everyone is arguing. AI can't hold that relationship. The PM still has to.
The PMs who will be hired in the next two years have those capabilities and can also move fast in AI-augmented workflows. The iterations are shorter. Decisions happen faster. The cycle time from insight to shipped test has compressed from weeks to days in teams doing this well.
What This Looks Like in Practice
In the agency and startup work I do across difrnt. and difrnt.ai, I'm watching this shift happen in real time. The people with the most impact aren't writing better product requirement documents. They're moving from identified problem to shipped experiment in days, not weeks, and they're using AI to close the gap between insight and execution.
The PMs who have been hiding behind process as a substitute for judgment are the most exposed. The ones who always wanted to move faster but were constrained by the old workflow are having their best year.
If you're a product manager reading this: what would you do if documentation, meeting coordination, and synthesis work took 80% less time? That question has a specific answer. The answer tells you whether you're building the right skill set or the wrong one.
The transition isn't about AI being better at product management. It's about AI eliminating the overhead that used to obscure who was actually doing the judgment-intensive work. That clarity is what's driving the restructuring.
FAQ
Why are product manager roles at risk from AI?
AI is automating the coordination and documentation-heavy work that traditionally occupied most of a PM's time: writing specs, synthesizing customer research, tracking competitive landscapes, and managing project timelines. What remains is judgment-based work that requires context, relationships, and decision-making under ambiguity. That work is valuable but requires far fewer people to execute well.
What skills should product managers build for the AI era?
Focus on judgment and speed. Can you identify the right problem faster than before? Can you move from insight to shipped test in days? Can you operate with less certainty and course-correct quickly? Those capabilities compound. Technical literacy with AI tools matters too, but judgment and execution speed are the differentiators that can't be automated away.
How should companies prepare for PM workforce restructuring?
Audit what your product managers actually spend time on. If the majority is coordination and documentation, your cost structure is vulnerable to restructuring. Companies repositioning early are redefining PM roles around ownership and outcomes rather than process management. Start by measuring what percentage of PM time produces direct product impact versus administrative overhead.