Ben Gregory at Kasava has an observation: product managers are weirdly well-suited for AI work. While engineers get frustrated when the same prompt gives different results, PMs have spent their entire careers dealing with outputs that never match the original spec. That comfort with uncertainty is why PMs are getting real value from AI tools. The data backs this up. Lenny Rachitsky's survey of 1,750 professionals found that 21% of engineers say AI makes their work quality worse. That's the highest dissatisfaction rate of any role. Meanwhile, over 70% of PMs and founders report quality improvements. Engineers expect precision. Code generation either compiles or it doesn't. PMs expect iteration. This is spawning a new role: product engineers. These are people who use AI to build what they used to delegate. LinkedIn shows a 45% jump in product engineer job postings. Companies like Vercel, Stripe, and Airbnb are creating dedicated positions. The market is putting its money where its mouth is. Salary data from levels.fyi shows product engineers command 15-25% premiums over traditional engineering roles. Gregory points to mode-switching as the real advantage. PMs shift between constraint thinking for technical specs and narrative thinking for roadmaps. They flip into empathy mode for customer research. Good AI prompting patterns demand that same flexibility. Same tool, different cognitive approach each time. People who prompt the same way regardless of context get mediocre results and blame the model. What does this mean for hiring? Companies might start valuing PMs who can build over engineers who can't handle ambiguity. For PMs, the career path is getting clearer: learn to ship, or watch someone else do it faster with AI.