Where to draw the line between Imitation and Inspiration?

I agree with every word of Mitchell Hashimoto’s essay on taste. I also can’t do anything with it.

His argument: AI has commoditized the part that used to be hard — execution — so taste is now the differentiator. His definition is clean: “taste is the ability to consistently make high-quality qualitative judgments where no objective metric exists.” Hard to build, and trivial to copy once it exists. The skeptics say the copying proves taste is fake; he says the copying proves the opposite — you needed someone with taste to copy from.

All true. All inert. “Taste” sits in the same drawer as “vision” and “good judgment,” the words you reach for after the fact to explain why one team won. I hate concepts that get lost in translation between the person who coined them and the person who has to act on them. So let me make it operational, in a language most of us already speak.

Imitation vs Inspiration

Every designer knows this distinction cold. It’s the cleanest line in the language, and it maps to something programmers know even better.

Imitation is copy by value. You duplicate the output — the exact UI components, the layout, the illustration. What you get is a snapshot, captured at the time of copying. The moment the original changes, your copy is stale. You’re chasing a moving target with a photograph of where it used to be.

Inspiration is copy by reference. You don’t duplicate the output; you internalize the principles, the why behind the decision, the judgment that produced the thing. Now you point at the live thing and stay in sync. Better: you can produce the next decision yourself, including ones the original hasn’t made yet.

That’s the operational version of Mitchell’s insight:

Tasteful work is easy to copy — but only after it’s been made.

Imitate the results without understanding the decisions that produced them, and all you’ll ever ship is what already exists.

The position isn’t worth defending anymore

This is the part of Hashimoto’s argument I’d push harder on. He’s right that production is being commoditized faster than taste, and that one person with a defined vision can now build what used to take a team.

Production got cheap partly because the companies selling you the tools are losing money on every token you spend. But cheap is cheap, and the subsidy doesn’t pick favorites. Your competitor’s clone runs on the same discounted execution you do. Which means execution is no longer where the gap lives. Any UI is a weekend. Any feature set is a prompt. The position — the thing you’ve built and can point to today — is the most copyable asset you own.

So if your moat is the artifact, you don’t have a moat. You have a head start, measured in weekends.

The moat is the derivative

Here’s the turn. AI commoditized the position, so the only defensible thing left is the rate: the velocity and consistency of your next hundred judgment calls.

Watch how the clone actually plays out in time. A competitor copies your current product. It takes them a weekend — generous. By the time they ship it, you’ve integrated another round of feedback, reversed two of your own flawed decisions, and made three new ones they haven’t seen yet.

What they ship is a photograph of where you used to be.

Hashimoto puts the value of taste precisely: it matters “exactly because it defines what everyone else chooses to copy.” The cloners are downstream of you. They can only copy decisions you’ve already made, which means they’re sorting through your past while you’re deciding your future. The asymmetry isn’t that they can’t catch up. It’s that catching up is the only move the position allows them — and the position keeps moving.

That’s what a founder should take from all this. Stop guarding the result of your work like it’s the asset. Spend the effort on the thing that generates results: the reasons, the constraints, the judgment that decides what’s worth shipping next.

Building from inspiration is how everything real gets made; nobody invents in a vacuum. And there’s a clean test for which side of the line you’re on: explain why each choice is right without pointing at the thing you copied. If you can, you internalized the reference. If your only answer is “because that’s how they did it,” you copied the value.

Draw the line there. Once you’re on the right side of it, the cloners stop being a threat. They’ll show up the moment your work is worth taking, and they’ll take everything they can see.

They own a photograph. You own the camera.