Why half of product managers are becoming obsolete — and what to do about it

Executive overview

The product management role is splitting in two: builders who love making things are thriving, while information movers — those whose value was coordinating, synthesising, and relaying decisions — are being made redundant by AI. This is not gradual; the skills that earned promotions three years ago are now liabilities.

The core insight: builders will have the time of their lives, but if you don't love building, you're in trouble — and roughly half of current PMs fall into that second camp.

The next 12–24 months will bring massive layoffs followed by targeted rehiring of AI-first talent. The path through is crossing a psychological threshold from fear into joy — finding one personal moment of building something that makes the new way of working feel worthwhile.

The state of the industry

• The dominant PM activity three years ago was moving information — relaying context up and down the chain. That role is being automated out of existence.
• The best PMs right now are earning more than ever, seeing more offers, and moving into founder and C-suite roles outside product.
• Mid-career PMs (30s) are disproportionately stressed: peak career demands collide with family, aging parents, and health for the first time.
• Companies are auditing whether headcount growth delivered proportional output — and concluding it didn't.
• Expect companies to shed large numbers and rehire at a fraction of the size, but exclusively AI-first profiles.
• The most open PM roles globally in three-plus years currently exist — builders are being hired aggressively.

What judgment means (and why it's the only skill that scales)

• Judgment is the ability to decide whether a change is good or bad, which of many possible directions to pursue, and whether something is worth building and releasing.
• As the cost of testing drops to near zero, 10–100x more changes will be presented for evaluation — judgment becomes the bottleneck, not execution.
• Information aggregation (status reports, standups, PRD writing) will be fully automated within two years at most companies.
• The job of a senior PM is increasingly: set direction, evaluate output, and build internal tooling to automate the product operating system itself.
• Companies are already fully automating product reviews and standups; the PM's job is to build the tools that make that possible.

Why builders are winning

• Builders have a direct connection between their ideas and customer outcomes for the first time — no waiting for designers, backlogs, or sign-off chains.
• The cost of building has collapsed; PMs can now prototype, test, and ship without depending on a team.
• Founders are emerging from senior PM ranks at a higher rate than ever: 14 of 125 heads of product in one community founded companies in the last 12 months.
• Product skills are moving into adjacent functions — one senior PM was recruited as a CHRO because companies want the obsolescence and judgment mindset applied to HR.
• The "builders wanted" label is now cross-functional: engineers, designers, and marketers who build are all competing for the same roles.

Why information movers are at risk

• Roughly half of current PMs built their identity and skill around communication, team-building, and information coordination — not building.
• That profile was well-compensated and promotable in the ZIRP era; it is now being actively de-prioritised in hiring.
• Saying "I'm not really into tech" is no longer a neutral statement about work style — it's a signal of misalignment with where the industry is going.
• This group may need to transition out of tech entirely, or apply AI tools to build new businesses outside the industry.

The psychological barrier to reinvention

• Humans are trained to find a stable system and minimise change — the skills that made someone excellent at the old game actively resist learning the new one.
• The better you were at mastering the previous model, the harder reinvention feels; there's no internal incentive to change when the current approach still works locally.
• Mid-career time constraints make reinvention feel impossible: the "equally disappoint everyone" model leaves no slack for learning.
• The target keeps moving — a one-week upskilling sprint becomes stale within three months.
• The block is psychological before it is practical; crossing the threshold is the prerequisite to everything else.

How to cross the threshold

• Find the first moment of personal joy with the new tools — an app built for a partner, a chief-of-staff agent, lights controlled via a script. This moment is when the shift happens.
• Joy is the antidote to burnout; once it's present, the human psyche creates time and energy that seemed unavailable before.
• Adopt the engineer's definition of excellence: obsolete yourself from everything you do — use AI to automate whatever you'd rather not be doing manually.
• Increase pace deliberately; treat the next two years like year one of a new job or relationship — bring the energy, find the reserve.
• Swallow ego: willingness to take something smaller now to stay current is the correct long-term bet.
• Stay long-term focused — the "skip" mindset is always optimising for the move after next, not the immediate title.

Practical AI stack and building approach

• Standardise on one tool rather than context-switching between many; consistent depth beats breadth.
• Start by automating one thing you currently do manually — matching people in a community, tracking open roles, summarising inbound questions.
• You do not need engineering skills to get value; you need to be opinionated about what good looks like and persistent about iteration.
• The trigger for catching the bug: the first time an agent runs overnight and delivers what you used to do yourself.
• Build internal tools first — automating your own product operating system before customer-facing features is where the near-term leverage is highest.

Where things are heading (2–3 year view)

• The chaotic transition period will stabilise; there will be a new routine to the AI-native way of working, just as internet-era product management eventually standardised.
• Brand and company logo matter less in hiring; interviewers now ask about tools used, live judgment demonstrated, and how you think — not what you shipped five years ago.
• Long tenures at large non-AI-forward companies can now hurt candidates, because what was learned there is diverging from current practice.
• Geography and diversity will take a step back as the AI wave concentrates hiring in the Bay Area and in profiles that match current team composition.
• Women in power years face a disproportionate burden: the expectation of nights-and-weekends building is structurally harder to meet.
• Engineers are converging on PM skills as coding becomes automated — the distinction between the roles will blur further.
• Designers face a split between pixel production (at risk) and taste and case-making (durable).
• PMs are positioned to become change agents across industries — carrying the "obsolescence mindset" into marketing, HR, education, and non-tech companies that need to modernise.

What the best people are doing right now

• They're attending hands-on show-and-tells, not slide reviews — sharing what they built, one-upping each other on agent capabilities.
• They're building chief-of-staff apps to manage their own inboxes and workflows.
• They're using AI-powered matching and recruiting tools for their communities.
• They're running agents trained on their own content to answer questions at scale, then reviewing the answers to find where they disagree with the model.
• They're vibe coding alongside TV — treating building as leisure, not work.
• They describe their experience as "smiling exhaustion" rather than plain exhaustion.