An MBA candidate finally voiced what everyone in the room was thinking: "Am I an idiot for investing in a PM career when AI is just going to raze the whole field?" Fair question. The answer is no — but understanding why requires getting clear on what the job actually is.
Product Management, at its core, is a few non-negotiable human elements: accountability without authority (you drive outcomes by influence, not command), context switching (every stakeholder group uses different vocabulary and trades in different currency), and the soft skills that are actually the hard skills — curiosity, empathy, storytelling, perseverance, situational awareness.
AI struggles with all of this. The craft involves combining these traits to coax collaboration and drive outcomes. The human experiences of connection are the firewall protecting the PM career.
The fear that AI will replace PMs is rooted in the fallacy that PM value comes from labor — writing PRDs, formatting slides, combing through Jira tickets. That's not the job. The job is establishing the what and why of a product and aligning everyone to deliver. Most of the labor is artifact creation in service of that — and AI radically reduces the cost of that work.
The cost of curiosity collapses to near zero. In the past, asking "what would have to be true for this to work?" started a weeks-long research project. Today, it's a conversation. Ask the question, get a list of necessary conditions instantly, and transform constraints from brick walls into variables you can solve for.
The old way: spend hours (days? weeks?) writing one PRD, agonizing over formatting so engineers don't push back. The new way: stream-of-consciousness into an LLM, ask "what would have to be true for this to work?", and use the reclaimed time to talk to actual customers about whether those conditions are realistic.
The modern PM keeps the North Star in sight but uses AI to generate the degrees of freedom needed to get there faster.
The modern PM is a synthesizer of context who uses human judgment to drive outcomes across diverse stakeholder groups. We're moving from creating from scratch to architecting options. AI automates the tasks of Product Management. It cannot automate the craft.
If we lose the human ability to collaborate and connect to do big things, we have lost a lot more than just a job.