Output isn't Design

Happy Friday!

I’m Peter, and I write about the intersection of AI, UX, and these wild times of change. The kids are wrapping up their last weeks of school/university, so it’s all endings and beginnings over here. And that’s what I wanted to talk about today.

There’s a long-standing discussion in design around not being a production body-shop. Nobody wants to just produce screens all day. And yet, that’s what a lot of design organizations end up doing, for much/most of their time.

At the same time, the same conversations have been happening in technology (“nobody wants to be a code monkey”), in product (“output versus outcomes”), in law, you name it. It’s not at all unique to the UX discipline.

And so I do not think it’s the useful discussion to be having.

The Output/Outcomes discussion is about time to think, and in a weird way, AI may give us that. The thing that is truly impacting UX and other disciplines right now is that Writing Code is Cheap Now. It’s still underappreciated by most people that are not engineers. That dramatic change impacts all of our processes, and does not replace the fact that every discipline needs Time to Think.

So the more useful discussion is:

1. What does code being cheap now mean for our discipline? How does it impact our assumptions, processes, deliverables?

2. What is next for us?

Every digital product discipline and title (information architecture, product management, content strategist, you name it) was created in the old world, when code was expensive. Should we let go of our identities? Do we still need the same “design process”?

My feeling is that it’s more useful to drop most of your assumptions and start building from scratch. Stop being a “UX designers” or “Information Architect”. That world is ending. The good stuff, the fundamental skills and insights you have, will come back and bubble up when they’re needed.

Note: this was inspired by Karri Saarinen from Linear (the Very Well Designed ticket tool), who wrote about Output isn’t Design: “The hard part of design is rarely generating the form. It is understanding the problem well enough to know what and how something should exist at all. There is use and place for these tools, but tools are not the design process.”

A few more interesting bits and pieces:

Michael Angeles shared a Claude skill to research a company’s UX before the interview.

Sean West on the value of interns: “Gone are the days of interns spending weeks to summarize a 400 page book so a principal can figure out whether to read it and which parts. These are the days of interns identifying, licensing and parsing gigantic specialist datasets in a few hours by spinning up their own apps.”

Dan Brown is using Claude for information architecture.

Brittany Hobbs is writing about using Claude for UX researchers.

These are wild times, but remember, you can do hard things, and you’ll be ok.

Health and happiness,
Peter
PS: I am working on new video series at model context experience, let me know what you’re trying to learn.