craft as ai moat/spring series 3 of 5
the bottleneck was never speed. it was judgment.
ai made the working prototype an afternoon's work, and the volume of mediocre software has never been higher. what is scarce now is someone with the taste and the standing to decide what deserves to exist.
productboard's hubert palan1, the company's founder and ceo, keeps making a point that matches what we see in the work: shipping fast is no longer the binding constraint. agents ship faster than ever. what most products are short of is someone deciding the right thing to ship.
for 2 decades, the product industry operated on a shared assumption: if we could just ship faster, we'd ship better. agile was supposed to solve it, then lean, then devops. each wave promised that speed would deliver quality.
then ai showed up and gave us speed. real speed: the kind where a founder with claude and a figma make subscription gets from idea to working prototype in a single afternoon.
the speed problem is, for practical purposes, solved. and a lot of the products still suck.
not all of them. but the volume of mediocre software has never been higher. the app stores are flooded with it, and so is linkedin. every landing page looks like every other landing page because they're all generated by systems trained on the same averages.
that volume is the evidence. the scarce input was judgment all along.
speed lets you iterate on a bad idea faster. it does not tell you whether the idea is worth iterating on, which of the 40 features to cut, or that the onboarding flow everyone thinks is fine is actually the reason retention is dying in week 2.
that call takes a different muscle: taste, experience, and the human layer ai has not touched.
the market is signaling this. figma's 2026 state of the designer report2 found that 45% of senior creative directors are now actively rejecting ai-generated assets for major campaigns. the output is often passable, and passable is exactly the problem. when everyone's output hits the same "good enough," the bar moves.
the same figma report caught another signal. designers now define craft as 2 things: visual polish (57%) and thoughtful problem-solving (47%). design systems, what most companies still treat as component libraries, are evolving into what figma calls "living frameworks that scale taste."
taste as infrastructure, rather than decoration.
the companies that win the next 5 years will not be the ones that ship fastest. they will be the ones with a strong enough point of view to know what to ship in the first place, and the craft discipline to make sure what they ship actually feels like something. direction is the scarce input. features and speed are the easy ones to buy.
sources