How AI is reshaping the product role | Oji and Ezinne Udezue

Oji and Ezinne Udezue argue PMs are now the bottleneck as AI accelerates engineering, and companies that rebuild codebases around LLMs will beat those sprinkling AI on legacy systems.

• Engineers are moving so fast with AI that PMs—not devs—are now the bottleneck; one company prototypes in 4 hours after a pitch call.
• Companies where the LLM becomes core to the codebase will beat companies that attach AI at the edge of legacy systems.
• Oji has written more code in the past year than the prior 10 years combined; he frames code as “architecture and English.”
• Top PM skills for 2025: curiosity, humility (“teachability = survivability”), high agency, eval-writing, and multi-model literacy—not just prompt engineering.
• A “sharp problem” is an old, persistent pain where a 3–10x improvement or 10x cost reduction triggers immediate willingness to pay.
• The “shipyard” team is six capabilities—PM, engineering, design, user research, data/ML, product marketing—working in near-continuous collaboration, not standups.
• Observational/ethnographic research beats AI-synthesized interview transcripts; what customers do matters more than what they say they want.

2025-09-07 · Watch on YouTube