Illustration labeled Creative Workshop, UX Design, Builder/Experience Lab, Experiment Build Play, Prototype, Iterate, Diversity, Innovation, Joy, Unity, and User Flow, with diverse people creating designs and technology.

Doing things that really matter

September 21, 2026

—

2–3 minutes

To read

On Saturday, during breakfast, the family had a great idea: they wanted a “Weasley clock” — a wall-mounted display inspired by a character from the Harry Potter universe that, instead of showing the time, shows the current location of each family member.

When we finished breakfast, I typed a quick description into ChatGPT on my phone and asked it to turn it into a prompt for building a prototype. We iterated on the spec a bit while I registered for Lovable, and then it took over and built the application. The whole process took less than five minutes.

The app has a fully functional backend, where I can set a typical day or connect to calendar feeds. It has a basic frontend dial and a simple interface for overriding the current schedule by tapping the hands. Perfect for iPad-on-a-wall interaction.

As I was demonstrating this new addition to the household — and trying to make sense of the ridiculous development process — I said: “It’s amazing how this thing just builds anything we come up with, and we get to focus on the things that really matter, like fonts and colors.”

Two different things matter here, and both are made possible by this AI development cycle:

Custom software for the long tail of needs and desires. An incredible number of new “use cases” will now be satisfied on a whim, and a lot of them will bring new kinds of joy to their users. We haven’t seen this kind of positive excitement since Geocities.
More diversity in UX design, and with it, more humanity in human-computer interfaces. This industry has been under siege since its inception — first by engineers (text interfaces, Microsoft), then by designers (Web 2.0), and lately by engineers again (design systems and SDKs). I can work in all these design languages, but my newfound joy is being able to do things tailored to my specific situation and imagination.

The last time we had this much fun in software design was when Kai’s Power Tools came out in the mid-nineties.

I’m calling on everyone to take this opportunity and have some fun with software in 2026. Soon enough, things will get complicated again — too expensive, too slow, something will go wrong, and we’ll become an engineering-driven industry once more.

But this year, we should all experiment.