Super Nice's CCO: AI made it easier to make something, not to make something people care about

AI can spin up a screen in minutes now. That means the thing you used to hire for, turning a vague brief into something that looks right, is no longer scarce. So the question lands on your desk: if anyone can make the artifact, what does your team still own? Let me catch you up.

The deep cut

• The artifact is cheap; judgment is not. When a tool ships a prototype in minutes, deciding what to build becomes the job, not the screen.
• Design earns trust, not applause. Obopay's payment screens had one job, to answer "is my money safe," and every AI product now faces the same test.
• Name who you are actually serving. Joshua Leigh splits "the designless" into three groups with different needs, and lumping them together weakens your call.

The thing that got cheap

Making a passable screen used to take skill, tools, and time. Now a prompt does a rough version. Joshua Leigh puts it plainly: a designer takes a client's half-formed intent and gives it form, and AI has not changed that job, only who can attempt it. The deliverable stopped being proof of anything.

That shifts where your team's value sits. A designer who spent years making pixel-perfect layouts on one platform is now competing with a text box. The skill that holds up is the one that decides what should exist at all, and what to cut.

Trust is the part AI can't fake

The designer who worked through four tech waves says every shift starts with the same problem: the tech works, but nobody trusts it yet. At Obopay, the hard part wasn't the interface, it was trust. Each confirmation screen had one job, reduce doubt. If anything felt unclear, people backed out.

AI products sit in that exact spot now. People don't doubt that the model produces output. They doubt whether they should rely on it. That gap is design work, not engineering work, and it does not photograph well for a case study.

Getting out of the way

The best design now looks like nothing. Every extra tap is a tax on trust, and users drift toward whatever feels easy. Checkout in two taps, onboarding that gets someone to one real win fast. The craft moved from decoration to removal.

There is a catch. Invisible does not mean opaque. When your interface starts deciding for the user, they still need to see what happened and undo it. Products lose trust when they strip control along with the friction, so the person feels something was done to them, not for them.

Know which "designless" you serve

The debate about AI and design gets muddy because people talk about one audience when there are three. Leigh names them: people who lack the access to hire design, the awareness to know what to ask for, or the motivation to bother. Each needs a different answer. A tool that helps a solo founder is not the same tool that helps a big team.

Some shops are answering by hiring senior humans, not automating. Super Nice's CCO put it well: AI made it easier to make something, not easier to make something people care about. That line is a hiring thesis. Judgment and taste are the scarce part now, and the market is paying for them.

Three questions for your team

• If a prompt can produce our screens, what is the specific decision only our team makes, and can we name it in one sentence?
• Which of the three "designless" groups are we actually building for, and does our roadmap match that one, not all three?
• Where in our product does the user feel something was done to them instead of for them, and what would it cost to give that control back?