PROCESS & CRITERIA
Where Does Judgment Belong in the Design Process?
August 2026 · 4 min read
AI isn't taking work away. It's moving where judgment happens.
This is something I've started to see in some projects as a Product Designer.
It is increasingly common to find solutions built with AI or low-code tools by junior or generalist profiles, with a very high delivery speed. In many cases, when those solutions reach review, senior judgment enters to correct what is already practically built. And there appears a difficult feeling to ignore:
The problem is not the speed at which we build, but the moment judgment enters the process.
The expert's work shifts toward the end: reviewing, correcting, and putting out fires, instead of participating in the decisions that define what should be built from the beginning.
I don't feel the problem is the use of AI. On the contrary, it's a huge opportunity. The problem appears when we accelerate execution without redesigning the decision flow.
I've lived moments where, when reviewing this type of solution, you have to go back to the basics:
- What problem are we really solving?
- Who are we doing it for?
- What hypotheses are we taking for granted without having validated them?
This leads me to an uncomfortable question: are we using technology to eliminate work… or to displace judgment?
Because they aren't the same thing. Eliminating operational work is positive. But displacing judgment generates systems where a lot is built, but little is decided.
Where does our value remain?
If tools allow more people to build functional solutions, where does the value of professional judgment in product design remain?
I don't have a closed answer, but I do have an intuition: value is not disappearing, it is moving toward the earlier phases of the process. Toward problem definition and the invisible decisions that sustain the final product.
Perhaps that is the most important change we are experiencing right now.
Is expert judgment entering earlier in your processes… or arriving when the product is already practically built?
