Back to Home Case Study

Finding my moat as designer in Ai era

Most AI tools today can generate a decent interface in minutes. Give it good context, write a clear enough prompt, and you'll get something that's "good enough" to ship. That's not a prediction anymore that's just where we are.

RoleProduct Design Lead
Timeline2026
PlatformWeb & Mobile
FocusStrategy & Research

Most AI tools today can generate a decent interface in minutes. Give it good context, write a clear enough prompt, and you'll get something that's "good enough" to ship. That's not a prediction anymore that's just where we are.

Defeated by speed

So where does that leave designers? Where's the exclusivity now? This isn't a rhetorical question for me. I've genuinely been sitting with it trying to figure out where I position myself, and what makes me stand out when the tools can do so much of the work I used to spend hours on.

Try it before you judge

The first thing I told myself was: don't be allergic to AI. Try it before you judge it. So I did  Figma MCP, Google Stitch, and just talking to Claude directly, feeding it context and asking it to generate HTML.

From that, I picked up two things that genuinely surprised me.

1st Findings

First, AI is really good at expanding an idea. Give it a seed of context, and it'll grow outward in directions I wouldn't have thought of. As humans, we tend to be deep in one domain and shallow everywhere else. AI doesn't have that limitation — it can show you how your domain connects to something adjacent that you didn't even know was relevant.

2nd Findings

Second, it's surprisingly good at restructuring how I communicate. My thinking isn't always linear. Sometimes the conclusion shows up in my head before the introduction does, and what comes out on paper is messy because of it. AI handles this seamlessly — it takes my scattered input and gives it back to me reordered, structured, viewed from a helicopter rather than from inside my own head.

But here's where it falls apart:

Problems

AI doesn't have taste. Most outputs feel the same. The layouts might differ, but the assets are generic — there's no identity in them. If you're working on a brand that needs a strong personality, you can't just lift AI-generated assets and call it done. The identity gets diluted, blended into whatever the model has seen a thousand times before.

Designer responding

So that's where I've decided to put my energy. I went back to my roots typography. AI can generate image assets. What it can't do, at least not yet, is generate assets and typography that work together as one cohesive system with intention behind it.

Maybe the moat isn't about being faster than AI. Maybe it's about depth in the places AI hasn't reached yet. Expanding my typography vocabulary isn't just leveling up a skill — it's a bet that taste, the kind that comes from years of paying attention to how letterforms and assets talk to each other, is still something only a human can bring to the table. If AI can match your speed, taste is what's left to win on.