Generated Form Is Not a Solved Problem

Most of the new AI design tools are built on a misunderstanding. They promise to generate interfaces faster, push words straight to product, or collapse design directly into code. The assumption underneath all of them is the same: design is the act of producing.

It is not. The hard part of design is rarely making the form. It is understanding the problem well enough to know what should exist at all, and how.

This is sharper in fintech than almost anywhere else, because half of what we are designing is invisible. We are designing trust. A merchant deciding whether to plug our checkout into their store is not evaluating the gradient on a button. They are reading a hundred small signals — how a confirmation state behaves, what a permission screen asks for, how an error message handles their money — and forming a judgment that no usability test fully captures. The output of good fintech design is often something the user never consciously notices. The output of bad fintech design is a support ticket, a chargeback, or a merchant who quietly stops sending volume.

Christopher Alexander named this in Notes on the Synthesis of Form. He describes design as the search for a good fit between a form and its context — context being the full set of forces that make a problem what it is. Human needs. Technical constraints. Conflicting requirements. Edge cases. Relationships you only notice after sitting with the problem for a while. In payments, that context includes things you cannot prompt your way to: regulatory reality, fraud surface, what a merchant has been burned by before, how a buyer in Manila interprets a redirect screen at 11pm.

You can already see what happens when teams skip that work. Products that look polished in a screenshot but unravel the moment money moves through them. Forms that ask for the wrong field at the wrong time. Confirmation flows that leave the user uncertain whether the transaction went through. The form is there. The fit — and the trust — is not.

This is why we advocate to deisgn visually over prompting. Working visually keeps us close to the problem. It is slow enough that it gives us time to think while we work. Moving things around, testing relationships, sitting with a flow until it feels honest — none of that is separate from the thinking. It is how clarity emerges, and in fintech, clarity is most of what trust is made of.

Writing works the same way. The act of writing forces you to organize what you think. Asking AI to write for you can produce text, but it rarely rearranges your thinking. Design has the same property. The value is not only in the output. It is in the understanding that comes through doing the work.

AI is still useful. Some work should be delegated to it. It can prototype, explore, surprise you. But that is not design. Design — especially when the outcome is something as intangible as trust — requires judgment, conversation, tension, and time.

The risk is mistaking generated form for solved problems. In fintech, the bill for that mistake comes due in money.