I recently shared a short post on LinkedIn about why I think it’s a great time to be a designer, and I wanted to expand on it a bit here.
Now that code is becoming a commodity, the bottleneck isn’t getting software to run anymore. Even with an agent writing the code, building a quality product takes time, dedication, and craft. Taste, vision, and obsession over details are what separate a sea of glorified AI demos from tools that actually feel great in the hand.
We’ve all seen plenty of those demos by now. Someone prompts an app over a weekend, records a smooth 30-second video for Twitter, and it looks impressive. Until you actually try to use it.
A demo only has to work once, on the happy path, for the person recording their screen. A real product has to hold up everywhere else. You have to think about empty and error states, how a transition feels when you interrupt it, whether the layout breaks on a small screen, or just all the small bits of feedback that make an interface feel responsive.
A coding agent can build all of those things easily once you ask for them. The hard part is knowing what to ask for, noticing when something feels slightly off, and caring enough to keep pushing past “it works.”
Redesigning this website was a good reminder of that. I built the new home page around an ASCII animation that morphs as you scroll through my projects. Since an agent was writing most of the code, getting a first working version took maybe an hour. Getting it to actually feel good took days of back and forth.
Almost none of that time was spent on pure engineering problems. It was just design iteration. I started with an ASCII version of my portrait before realizing it felt too static. I tried hijacking the scroll to snap between projects, hated how it felt, and went back to native scrolling. I spent hours tweaking the shape of the ASCII canvas, the easing of the transitions, and little details like the interactive pig in the header or subtle UI sounds. No single prompt gives you that, it just comes from using the thing, noticing what’s annoying, and fixing it.
What really changes for me as a designer is how fast I can go from an idea to testing it in the real medium. I used to spend so much time explaining how something should work: annotating Figma files, recording motion specs, and hoping nothing got lost in handoff. Now I can just prototype directly in the browser, feel it right away, and try three different ideas in the time it used to take to spec one.
When building software gets this cheap, simply shipping something that works stops being an advantage. The value shifts to deciding what is worth building and sweating the last 10% of polish that usually gets cut. And honestly, that’s the most fun part of the job anyway.