Why AI-built products feel unfinished

A team ships a first version in weeks instead of quarters, puts it in front of real users, and gets feedback nobody knows how to act on. Not “it is missing X”. Something vaguer: it feels rough. It feels like a demo. One customer says it looks like an internal tool.

The feedback is accurate, and it is not about features. It usually comes down to the same four things, and a generator has the least reason to produce any of them.

1. Nothing on the screen is more important than anything else

Generated layouts are even. Every card the same weight, every heading the same size, every button equally loud. That is not a style problem; it is the absence of a decision. Someone has to decide what the screen is for, then make one thing win. On an even screen the reader has to make that call, every time, and they call the effort “confusing” long before anyone can name the cause.

2. The empty and error states were never drawn

You prompt for the screen, so you get the screen: populated, correct, the happy path. But a new customer meets the empty version first, and an unlucky one meets the error. Those two states set expectations more than the populated view does, and in many AI-built products nobody designed them. They show the framework default, which is nothing.

3. The edge cases nobody thought to ask for

A generated interface handles the inputs you imagined. Real data is less considerate.

• The list with four hundred rows, and no way to find anything in it
• The name that is forty characters long and breaks the row it sits in
• The payment that fails at the wrong moment, in a flow with no way back
• The customer on a phone, on a train, on a connection that drops

None of these is hard to design for. They are hard to remember, and remembering is the first thing fast drafting skips.

4. There is no system underneath

This is the expensive one, and it compounds. Without tokens, components and rules, every new screen is a fresh negotiation: what spacing, which gray, does this button look like the one two screens back. For a while the answer is “close enough”. Then you have forty screens, six grays, and a team that argues about padding in code review.

A design system is not a style guide nobody reads. It is what lets an engineer build a screen no designer drew and have it come out right.

So is AI the problem?

No, and we would be hypocrites to say otherwise. We use the same tools every day: to explore twenty directions in the time three used to take, and to scaffold the code under a build. What they do not do is decide. They have no taste and no stake in whether the work lands.

Generating and finishing are two different jobs. Generating got far cheaper, and that is good news. Finishing did not, and pretending otherwise leaves a working product feeling like a prototype.

If your product feels that way, send it to us, however rough. We will tell you which of the four is costing you—and whether it is worth paying anyone to fix yet.