Skip to main content

Product Development

Design products and services where you have proven demand by understanding the jobs they need to do, and the outcomes that signify success.

Context

  • Products: Make it easier to make progress
  • AI Agents: Make it easier to use products

Roles vs Products

RoleProducts
Product Managerproductboard, aha, airfocus
Market Researcherexploding topics, trendhunter, sparktoro
Product Validatorvalidately, userfeed, prelaunchhero
Programmercursor.ai, replit, claude 3.5
UX Researchermaze, hotjar, usertesting
UX Designerv0, playground, galileo ai

The Agent-Native Shift

AI tools are changing what each role in product development actually does. Market researchers who used surveys now use AI to synthesize signal at scale. UX researchers who ran moderated sessions now instrument products and let usage data surface the insight. Programmers who wrote code now direct agents that write code.

The question for a product team is not "which tools are best for each role?" but "which roles still require a human in the loop, and why?" The answer changes the org structure, the hiring plan, and the product development velocity.

What doesn't change: The job of understanding demand before building supply. The tools change. The discipline of qualifying before quantifying stays constant.

Questions

Which product development role is most exposed to AI substitution — and which role becomes more valuable as AI takes on more of the execution?

  • At what stage of product development does AI assistance produce diminishing returns — where does human judgment still outperform agent output?
  • If product development is a feedback loop between customer signal and product response, which part of that loop is currently slowest — and what closes it?
  • How does the AI-native development stack change the minimum viable team size for shipping a product that generates real revenue?