Product design's future

Over the last year, I’ve been trying to figure out what’s going to happen to product design as a field.

I started my career as a graphic designer, and I loved it. But after a few years, I realized it wasn’t a great path to really build a life on. Namely, though I got a lot of enjoyment out of it and did some great work, I hadn’t cracked the nut of how to make a living.

So I transitioned into building websites, and eventually started hearing about this thing called product design, where I’d get to design software.

I learned all I could, started taking on any project that even remotely resembled that field, and then caught a break when someone hired me based on that scrabbled-together experience (thank you!). Now, years later, I’ve had time to learn and grow in this field, and I really love it.

That doesn’t mean I’ve learned everything there is to know. It’s not even close. Each day is a new adventure. But there’s a world of difference between now and even a few years ago.

All that said, with the advance of AI, this field has been changing fast.

Two years ago, product design held a great spot in the lineup of building software. We talked to users (or worked with UX researchers to do that alongside us), worked out questions with stakeholders, wrestled through concepts with product managers, then passed off designs to developers.

It was a tried-and-true process, and it mostly worked.

But even then, I struggled against it. When I transitioned out of graphic design, I learned to code my own websites. Granted, that was a much simpler time, and things have changed. But I learned how to translate designs into code.

For the early part of my time as a product designer, I mostly just created the designs and hoped and prayed they’d work out when things hit production.

That bugged the hell out of me.

So I found every way possible to skirt around that.

I started becoming best friends with the amazing engineers on the teams I joined, and began working more closely with them. I’d take the requirements, then sketch out what that could look like on my iPad. Then, in many cases, the incredibly generous developers would tag-team with me to take those sketches directly into code.

This wasn’t how product design was supposed to be done. And I bucked up against that unwritten rule—somewhere, someone thinks there’s a product design degree that defines each step you must take—all the time.

I didn’t want to create picture-perfect Figma files, because I knew, based on market evidence, that the reality of development was far trickier—and every second spent in Figma was a second wasted on getting something out to users.

So I employed every skill I could find to move from customer desires and stakeholder needs and translate that somehow directly to developers with any tool I could use. During those years, my iPads racked up thousands of hand-drawn sketches. I’m still quite proud of all those.

Then I tried to get a job at a company that only hired designers who code. They wanted a product designer to take an idea all the way from initial concept through to production front-end code. That excited me. It was getting back to where I started. But my coding skills had atrophied just enough that it didn’t make sense to bring me on.

I started dreaming again about recreating my first experience as a web designer: concept to code. I tried to learn to code, but my designer brain struggled with all the million dependencies that are now required just to get an environment up and running. Every time a developer walked me through it—and I took meticulous notes—my carefully crafted local environments would just crumble at the first error.

So I kept going, kept struggling, until sometime last year when a friend nearly forced me to try Claude Code.

Now, in 2026, I see the future.

Basecamp, that company I tried to get a job at years earlier, had it right. Shrink the number of people required to ship an idea, have two or three smart people work on something, own it together, and build without a dozen heads being forced to communicate over the radius of a button.

That’s where we’re at, and I love it.

And that’s why I think we’re entering a world where Figma (and Sketch, and InVision, and all the rest, R.I.P.) will disappear. Great designers will just take problems and get them into prototypes, then work with developers to make sure they’re built right.

That’s the future, and it’s here now. It’s just taking time for our industry to catch up.

If you’re someone who has been in product design for a long time and has all your processes flushed out, that’s fine. Don’t freak out and jump ship. Your process works for a reason. The reason I adopted Claude Code so readily is because I was already struggling with how design was supposed to work. My brain pushed back against the requirements of getting a perfect picture in Figma—one that would deteriorate at the first whiff of code. So I was ready to adopt something, anything, that improved on that process.

If you’ve already nailed that process that I disliked, you can take things a little more slowly. Play around with tools like Claude Code, and start to see how they can help you. We’re probably still a few years out, but at some point I see most of these design tools going away, or having to pivot hard.

That doesn’t mean we’re entering a world where someone can tell AI to build a cool app and walk away. Or if we are, I have no interest in being part of it. Rather, we have an opportunity to build really cool stuff, using LLMs as a piece of the puzzle to iterate, editorialize, and be the creative directors of our own projects.

If you’re someone who just loves a specific part of the design process, whether it’s UX or UI-focused, don’t despair. Keep doing that, and offset the parts you don’t like.

And most of all, reach out to someone else to learn together. If I’m available, I’d be happy to talk.