things are weird

I’ve been designing and building websites for what now feels like an eternity. And I don’t think there’s ever been a time quite like right now, where it almost feels like we’ve been teleported back in time and asked to relearn a lot of what we thought we knew.

The last few years have been this weird and wonderful trip. We’ve somehow gone from “that’s supposed to be Will Smith eating pasta?” to “holy shit, is that Will Smith eating pasta, or is this another AI video?”

And that happened really fucking fast.

Over the last few months in particular, there’s been what feels like another major shift in how we design and build things. We’re now seeing designers build entire websites using tools like Claude Code or Codex, without a traditional “design tool” like Figma involved at all.

I see a lot of concern from designers about what this new era might mean for their jobs. Or this growing feeling that if they don’t keep up with absolutely everything happening around them, they’re somehow going to fall behind.

I get it.

When the tools we use and the way we work seem to change 10x over such a short period of time, it can be difficult to process. It’s exciting, but unsettling at the same time.

You might try one of these tools and it doesn’t quite do what you expected. Maybe it completely misses the point. Maybe you spend more time trying to fix what it made than you would have spent just doing it yourself.

But then you try something else.

And suddenly, something unfolds right in front of you in a matter of seconds, almost exactly as you imagined it. I think that’s the part that can feel incredibly weird.

The gap between having an idea and actually seeing it come to life is getting smaller and smaller. It now feels strange to iterate and iterate and iterate in Figma on a single idea before ever seeing what that idea feels like in the browser.

I should say, that part of the process is still important. I still use Figma every day. And still spend an unhealthy amount of time moving things around by a few pixels until they feel right. It’s just that bringing those ideas to life is becoming a lot easier… and quicker.

For years, I was pretty intentional about how I structured my workflow around client work, or whatever project I happened to be working on. And for the best part of 10 years, that workflow stayed largely the same.

Sure, a new tool or feature would come along every now and then, but nothing ever really changed the fundamentals of how I worked.

In fact, I’d go as far as saying that most of us freelancers probably follow a pretty similar process. You speak to a client, understand the problem, explore some ideas, open Figma and eventually put that first pixel on the canvas. There’s a familiarity to it. You know what comes next.

One thing I’ve found helpful is deliberately stepping outside of the workflow I’ve spent years getting comfortable with.

For me, that meant trying Claude Code. It was foreign to me, but parts of it felt familiar. I found myself describing something I would normally have spent an hour laying out in Figma, only to watch it appear in the browser thirty seconds later.

It wasn’t perfect. But that almost didn’t matter.

Of course, I can still do all of this in Figma. I can move things around, change the type, explore different directions and iterate on an idea until it feels right. I’m not suggesting that suddenly goes away.

The difference is how quickly I can take that idea a little further. With some practice, something that would normally live as a static design on a canvas can become something I can actually use, click around and interact with.

Suddenly, I had something different in front of me to react to. I could look at it and think, no, that spacing feels wrong. That type needs to be bigger. That interaction feels weird. What happens if this moves over here? What if the entire thing works completely differently?

Everything about working this way initially felt alien to me. And as a designer with absolutely zero background in engineering or programming, some of it still does.

When an agent is building something and starts telling me it’s refactored this, installed that, created some component or changed something in a file I’ve never heard of, I’d be lying if I said I understood everything it was doing.

Most of the time, I don’t.

I’m sure understanding the technical side makes you better at using these tools. In fact, I’ve already found myself slowly picking things up just by building things with them. But I don’t think you need to sit down and learn how to become an engineer before you’re allowed to open Claude Code and make something.

For me, the more important skill right now is understanding what I actually want to do with these tools.

Because making something faster doesn’t necessarily mean making something better.

You still need to know when something feels wrong. You still need to make decisions. You still need taste, judgement, ideas and an understanding of what you’re actually trying to communicate.

Making is getting easier. Knowing what’s worth making hasn’t.

And I think that’s what I’m slowly starting to realise through all of this. I don’t need to have some perfectly defined new workflow. I don’t need to use every new tool that comes out. And I definitely don’t need to understand all of it.

For the first time in a long time, the workflow I’ve spent years getting comfortable with is being pulled apart and put back together again.

Rather than trying to hold onto exactly how I’ve always done things, I’m enjoying figuring out what the next version of that workflow looks like.

I have absolutely no idea what that’ll look like a year from now.

And I’m starting to think that’s a good thing.