craft as ai moat/spring series 1 of 5
when the floor rises, only the ceiling matters
a year of product design has flattened into the same gradients and the same card layouts. when the floor rises for everyone, the ceiling is the only thing left to compete on.
merriam-webster1 named "slop" its 2025 word of the year. a word that started as insider shorthand for low-quality ai output is now the thing webster's editors picked to define the year. culture has seen enough of it.
the last 12 months are the most visually homogeneous stretch of product design we can remember: gradient blobs, card layouts, and color palettes that feel picked by committee, because in a sense they were. a committee of training data, optimizing for the average of everything it had ever seen.
the tells are getting predictable. emoji standing in for iconography. gradient circles with letters where profile photos should be. ai-selected palettes that have no harmony because no human ever sat with them. layouts that look fine in a screenshot but fall apart the second real data shows up: a billing tier, a long customer name, a rich analytics dashboard. you feel the product as assembled rather than built before you can put a name on it.
averages are unremarkable by definition. if ai generates output by finding the center of mass of its training data, that center is the middle of the distribution, not the edge where the interesting work lives.
in 2024, generating anything felt like magic. then the floor kept rising. now everyone has the same tools, and everyone's baseline is the same baseline. competition shifts to the ceiling, and the ceiling is craft.
craft is the thing you cannot prompt into existence. transition timing that makes something feel expensive. spacing that is intentional because someone chose it. copy that sounds like a person wrote it because a person did. edge cases handled the way someone who actually uses their own product would handle them. that work is not decoration sitting on top of the product. in 2026 it is most of what is left to compete on.
the builders who figured this out early are already pulling away. the ones still optimizing for output speed are producing more of the same, only faster, which is the opposite of what the moment rewards. the floor rose. the ceiling did not move. the gap between them is the work.