AI Is Helping Design Finally Get Real

Working closely with clients, sketching things together — these are good things. The core social task of learning how people think, helping them visualize solutions, and collaborating is not going away just yet.

But let’s face it. White-boarding, paper, static mocks, high fidelity Figma prototypes, and literally all the related processes never overcame the fundamental flaw: it’s not real, not fully functional.

The problems that arise are fundamental:

Since it’s not real, people have to “imagine” — you have to disrupt user testing and pitches to explain why it’s not real and why you can’t click this or that. You have to trust that people will understand and hope that when you do actually build it, the idea doesn’t blow up.

You still try though. You try to find creative ways to test. Complex things take a long time to mock up and update, and in the end it still doesn’t fully work. And that makes it hard to see relationships between things, hard to see business rules and real data in action.

You rely on words to fill the gap but words invariably let you down — rules take too long to document, you get miscommunication, long threads of discussion, missed design annotations, and so forth.

With AI, you don’t have to approximate anymore. You don’t have to work through 5 levels of fidelity. You can build a version and test it, and that is now feasible. What that helps improve is the fundamental quality of the evaluation experience.

Whereas product folks used to say “Ship it, break things”, now you can ship it (make it real) without breaking anything.

#Design #AI #ProductDevelopment #UX #Innovation