Taste
What would you remove?
As AI makes production cheap, curation becomes the bottleneck. Anyone can generate a thousand options. Almost nobody can pick the right one. Taste is pattern recognition applied to quality — knowing not just what works, but what's worth doing.
| Without Taste | With Taste |
|---|---|
| Ship everything | Ship what matters |
| Default to more | Discipline to subtract |
| "It works" | "It's right" |
| Quantity as strategy | Quality as moat |
Subtraction Protocol
Good taste is knowing what to remove:
| Test | How to Apply |
|---|---|
| Does it earn its place? | If you removed this element, would anyone notice? If not, remove it. |
| Does it serve the goal? | "I like it" is preference. "It serves the goal" is taste. |
| Would adding more improve it? | If the answer is "maybe," it won't. Subtract. |
| Is it finished or just full? | A piece is finished when there's nothing left to take away — not when there's nothing left to add. |
The best work has negative space. The silence between notes. The white space on the page.
Standards vs Opinions
| Opinion | Standard |
|---|---|
| "I don't like it" | "It doesn't serve the goal because..." |
| "This feels off" | "The contrast ratio is below 4.5:1" |
| "I prefer the other one" | "This version has three competing focal points" |
| "It needs something" | "The hierarchy doesn't guide the eye" |
Taste becomes useful when you can articulate why something works or doesn't. Without reasons, you have preferences. With reasons, you have judgment.
Building a Reference Library
Taste develops through exposure to excellence:
| Practice | What It Trains |
|---|---|
| Study the best in your domain | Calibrate what "good" looks like |
| Study the best outside your domain | Import standards from adjacent fields |
| Articulate specifically what's good | Move from "I like it" to "here's why it works" |
| Collect examples of excellence | Build a library you can reference and compare against |
| Compare your work to the reference | Notice the gap honestly. Then close it. |
Don't just consume. Analyse. "This headline works because it promises a specific benefit." "This interface works because there's one clear action."
Micro-Judgment
The small decisions compound into the whole:
| Domain | Micro-Decision | What Taste Looks Like |
|---|---|---|
| Writing | Word choice, sentence rhythm | "Decided" instead of "made a decision" |
| Design | Spacing, alignment, type size | 4px of padding that makes everything breathe |
| Product | Feature scope, default settings | Knowing what NOT to build |
| Strategy | Which opportunity to pursue | Saying no to good so you can say yes to great |
| Hiring | Who to bring in | The person who makes the team better, not just bigger |
The Shadow
Paralysis. Never shipping because nothing meets your standard. Snobbery that mistakes exclusion for excellence. Taste without execution is just criticism. Ship, then improve.
By Archetype
| Archetype | Taste Style |
|---|---|
| Realist | Judges against evidence, not feeling |
| Dreamer | Knows what excellence looks like — the vision |
| Philosopher | Judges against principles — does it align? |
Context
- Critical Thinking — Evaluating claims; taste evaluates quality
- Orchestration — Directing AI output requires taste to judge it
- Pattern Recognition — Taste is patterns applied to quality
- Positioning — Taste shapes what you're known for