The PM is dead. Long live the builder.

Both things are true. They're the same sentence.

Cursor and Claude Code write the code. v0 and Lovable build the prototype. Granola runs the customer calls and synthesizes them before you finish your coffee. PostHog AI answers the retention question in plain English.

A founder with taste and a laptop can ship more in a weekend than a six-person product team shipped in a sprint two years ago. That's the good news. The bad news is the same news. The role built to coordinate that six-person team has nothing left to coordinate.

What PMs did

1. Talk to customers and figure out what to build
2. Write the spec
3. Work with design
4. Work with engineering to ship it
5. Read the data and decide what's next

Everyone has a faster, cheaper substitute now. Granola does the synthesis. ChatPRD turns a voice memo into a PRD that a model will actually follow. Lovable builds the prototype in an afternoon. Claude Code writes the code, and you review the diff. PostHog AI runs the analysis. The PM was the wire coordinating those functions, but that coordination layer is no longer needed. The market is already pricing this in.

Airbnb folded product management into product marketing. Microsoft cut PMs at 3x their share of headcount in 2025. Shopify made it formal: prove AI can't do the job before you ask for headcount. Midjourney is ~$500M ARR with ~40 people. None of them paused to hire a Director of Product on the way up.

Founders, you don't need a PM

You need the tools.

Three customer calls on Monday. Granola synthesizes them. Draft the PRD in a Claude conversation on Tuesday. Prototype in Lovable on Wednesday. Review the diff Claude Code wrote on Thursday. Ship Friday. Ask PostHog what happened Saturday. That week used to take a six-person team a full sprint.

The reason founders historically handed product off wasn't strategy. It was a capability. Founders couldn't write SQL, couldn't make design files, couldn't ship code. All three are a sentence away now.

What's left for PMs

Three real doors. Three closed.

Real:

• Founder. You sat next to one for years. You absorbed the playbook. Go.
• Product engineer. Ship code yourself. One person, end-to-end, no handoffs. The wave comes for product engineers, too, just later. Use that runway to get to founder or GM, which are the roles that survive once the building is commoditized.
• Domain specialist. Healthcare, defense, fintech compliance. Trust that AI doesn't replace. Narrow door. If you don't already have the domain, you won't get it in 18 months.

Closed:

• Strategic PM. The strategy was the founder's. You were executing it.
• Growth PM. PostHog and Amplitude run the experiments. You ran the meeting about the experiment.
• Platform PM at a mid-size company. Pure coordination. First to go.

Senior PMs: become GMs

The most successful PMs I know aren't PMs anymore. They left to start companies. The ones who stayed inside took on a P&L and now coordinate teams that do more than just write PRDs. They own a marketing budget, support headcount, and a revenue line. They make hiring decisions across functions. They argue with the CFO about margin, not with the design lead about button placement.

That's the pattern. Stop calling yourself a product manager. Run the product like a business.

A GM owns a P&L. They split capital three ways. Build the next thing, innovation. Get customers to the thing, growth. Keep the customers you have, retention. That role doesn't go away. It gets more valuable. Once building is cheap, the constraint moves to "are we building the right thing, for the right cohort, at the right margin." That's a business question, not a product question.

A senior PM with ten years of pattern recognition is most of the way there. The gap is marketing, support, and a real number on the wall. Go get them. Take a P&L. Tie your comp to it.

Junior and mid PMs

The ladder you climbed onto isn't getting rebuilt. APM to Director made sense when coordination was scarce. Coordination isn't scarce anymore.

If you're junior, pivot now while it's cheap. Become a product engineer. Learn to ship code with Cursor and Claude Code. Or go be employee three at a real startup, get into the room where the company gets built. Title doesn't matter. The room does.

If you're mid-level, you're in the squeeze. Three options:

1. Take the founder swing. You have the pattern recognition. The next 24 months are the cheapest time in history to start.
2. Find a company that'll pay you to learn a regulated domain. Healthcare, fintech, defense. Build depth that compounds.
3. Push for an end-to-end scope at your current stage. Stop running standups. Start shipping. If they won't let you, that's the signal.

The bottom line

PM was never a craft. It was a position on an org chart. Crafts survive technology shifts. Carpenters still build. Writers still write. Surgeons still cut. Designers will still create high-fidelity UI. Engineers will still write code. Positions don't survive because positions exist to solve a coordination problem at a specific moment in time. When the coordination problem dissolves, the position goes with it.

Learn more at Uristocrat Studios