Let me be honest with you. When I first started using AI tools in my workflow, I felt a bit like I was cheating. Twenty years of doing things a certain way — building moodboards by hand, hunting references for hours, filling sketchbooks before touching a screen — and suddenly there's a shortcut. It took me a while to make peace with it.

But here's what I've figured out after actually using these tools on real client work: the shortcut isn't in the design. It's in the thinking.

Let me explain what I mean.

When I start a new branding project, the hardest part isn't drawing the logo. It's finding the right direction. It's understanding what the brand should feel like before I can make it look like anything. Historically, that process involved a lot of time — client interviews, competitive research, mood boarding, internal back-and-forth. Days of it, sometimes. And most of that time was me trying to translate a brief into a visual instinct.

Weavy cut that process in half. Not by doing the design for me — but by letting me test directions at the speed of thought.

Now I spend the first hour of any project in Weavy. I'm not trying to generate final assets. I'm generating atmosphere. I'll write a prompt like "the quietness of a northern European kitchen, pale light, considered objects" for a minimal wellness brand, or "humid city at 2am, neon reflection on wet concrete" for something more charged. I generate 40 or 50 images and move fast — yes, no, yes, no. The yes pile tells me something I couldn't have articulated from the brief alone.

That yes pile becomes a reference board in Figma. I pull in the images that resonated, add real-world references alongside them — typefaces I'm considering, color systems I respect — and suddenly the mood board has logic. There's a direction I can commit to.

Then Weavy closes and Figma opens, and everything from that point is made by hand. The way it always was.

Claude plays a different role. When I'm working with a client who doesn't have a copywriter — which is most startups — I use Claude to draft positioning statements, website copy, and brand voice guidelines. It's like having a very patient collaborator who doesn't mind being wrong three times before we land on something good. I give it the brief, the target audience, the feeling we're going for, and it gets me 70% of the way there in ten minutes. The rest is editing.

What hasn't changed, and what I think people misunderstand about AI in design: the judgment is still entirely human. Knowing which direction is right for this brand, in this market, at this moment — that's not something you can prompt your way to. It's pattern recognition built over years of doing the work. AI didn't replace that. It gave me more time to use it.

The designers I worry about aren't the ones using AI. They're the ones who think that because they can generate beautiful images quickly, they don't need to develop the underlying judgment. The tool amplifies what you already have. If what you have is good, it gets better. If what you have is thin, the cracks show faster.

Use it. But know what it is.