In the past month I've spoken with three former colleagues who were software developers in one way or another who were planning on a career pivot of one type or another. Some of them were looking to go into the trades, others were looking to retire early and take a service job just to get out of the house. All of them had the same fundamental reason for wanting to leave: building software in the age of LLMs is deeply unsatisfying to them. They miss the craft of software engineering: remembering libraries, thinking through algorithms, weighing approaches to problem solving, considering the architecture of a public facing API, creating documentation. All of these carefully honed skills and acquired knowledge is suddenly almost impractical and worthless, it almost overnight became a quaint sort of hobbyist craft. Artisanal code, how cute. In its place is prompting LLMs, documenting for LLMs, and thinking how to have LLMs evaluate the quality of the output of other LLMs. Not only was it not what they had signed up for, it wasn't even the same job. It's as if we took a master woodworker and said "you're good at working with a plane and a chisel, let's have you manage this furniture factory. No one really knows how the factory works and it's on the other side of the world but it generates chairs and tables, so it's basically like what you used to do." These friends and former colleagues all want out, to get away from this way of working and thinking.
I've always been in between writing code and building hardware and doing design. Really, my value was always that I could do enough of both that I could be the translator between the disciplines. That sometimes made folks who were more squarely in the world of designing think that I had some sort of superpower. But a few months ago I went to an event at Teague where a friend who works in design asked me: "what does AI mean for designing?" I said something like: "you're now a lot more valuable than me."
I think it's a bad time to love the craft of building software. Conversely, I think it's a great time to love thinking about software. What it should do, who it's for, why we have it, what it's supposed to do for us. You can think of an idea and have that idea in front of real humans almost instantly. You can test and explore and iterate and improve almost instantly. The days of banging your head against code to build features, wasting days on small bugs, the pains of making software, these may be very nearly over. They aren't mostly over just for software engineers, they may in fact be mostly over for everyone. This means that many people who were involved in the process of making software can now simply make software. Of these people, the ones that I'm most interested in are designers. Designers tended to be the ones who were thinking most deeply about what the software should do for the people who use it. Not what the software does, but what it does for people. The design process has always been so closely intertwined with making software but design itself has always been constrained by the process of building. That is less true now. And that is why I think that now is, in some senses, a wonderful time to be a designer and to do design work. Of the two careers I've had, I'm quite sure that it is far better to be a designer now than it is to be a software engineer.
This is not to say that there are not challenges to design as a way to make a living and to designers themselves; there most certainly are. Figma, perhaps the most commonly used tool in digital design, is automating away a great deal of the craft of using that software itself. Instead of thinking through visual or interaction details, learning the tool and then learning how to think with it, you simply prompt an LLM for dozens of versions and pick the one you like the most. In many cases the texture, the actual practice of designing, the day to day work, has changed dramatically. Instead of drawing, arranging, clicking, designers are prompting. For rote UX design like website layout that may be fine, to create an initial prototype or quickly update a part of an interface. For the nuanced work that makes design both meaningful and valuable, it's dramatically different. The best designers that I have worked with, and I have been lucky enough to work with truly great ones, don't think in words or two sentence descriptions. They don't want to run a black-box factory. I'm far enough from their level of skill that I would never pretend that I know how they think or work. But I know that we are losing a way of thinking about this kind of work. That does cheapen their skills and the craft that they hone and I know that that hurts. I've too have spent years learning skills that are now meaningless for tools that no longer even exist. However, this loss of craft is also stripping away something tangential to the actual work itself. The work is to think about what humans want and need and how tools and technologies fit into that. That's the constant across all kinds of design, industrial to digital to service. That is what design does. Design isn't simply "Figma user" or "CAD expert". To be fair, software isn't entirely encompassed by "C++ typist" but it is much closer to that than design is to "Photoshop user". One job is much more tied to the ability to use the tool. Designers aren't supposed to use the tool in order to use the tool, they're supposed use the tool to think about "what it does".
I've heard designers worry about the decline of the influence of design at the top levels of companies, that the Chief Creative Officer role rarely, if ever, exists any more. They're right to worry but that doesn't mean that designing itself is no longer important, it means that corporate hierarchies may not recognize that importance. There is a difference between whether doing design work is a way to make a great career and a way to make good things for humans. I can't pretend that I know whether design will stay a great career or even be a good job. If I had that kind of insight, I would have retired wealthy long ago. But I do know that doing the work of design now is far more valuable than it has ever been. We can find much to bemoan in what we lose but it's worth appreciating what we are gaining and what that might mean for what we've always been supposedly trying to do.