Taste is not it
Why we think "taste" is overrated as an explanation in design, and what we check options against instead, especially with AI tools.
- Note
- June 2026
- 1 min read
“Taste” gets used a lot in tech right now as the reason some products are good. We think it’s overrated as an explanation.
In February, Joel replied to a post making fun of people in tech who say they have “taste”, as he put it on X: “At this point, taste is totally overrated and it’s definitely not it.” A few minutes later he posted it on its own: “Taste is not it.”
What we lean on instead
We don’t have a big theory about this. Day to day, what helps us is more ordinary:
- Knowing who it’s for and what it needs to do. Options get checked against that.
- Feedback that names the problem. “I don’t like it” doesn’t help anyone fix anything. “The spacing is fighting the type” or “people can’t find the Save button” does.
- Watching people use it. A prototype in front of real users settles more than a debate in a review.
All of these can be explained to a client and written down for a team, or for the AI tools the team builds with.
Where AI comes in
AI tools make it easy to generate a lot of options, and the usual line is that the human part is “taste”: picking the nice one. We’d rather check the options against what the product needs to do, because that can be written down and shared.
If you want a design partner who explains decisions in plain words, send your brief.
Asked & Answered
We think it’s overrated as an explanation. Specific reasons, like who the product is for and what it needs to do, are more useful.
Feedback that names the problem, like “the spacing is fighting the type”, instead of “I don’t like it”.
Keep reading
- The last ten percentThe difference between good and lasting lives in the finish. It is where humans still win.
- Form follows intent"Form follows function" is half a sentence. The other half is the one that lasts.
- Why every AI startup looks the same, and how to brand one that doesn’tPurple gradients, sparkles and a chat box: why AI startup brands converge, and how to build an AI brand people can tell apart and actually trust.