Jarrett Fuller

My latest essay for Fast Company was published at the beginning of the month while I was on vacation, so I’ve been remiss in posting about it here. It’s called Design is on its deathbed. Long Live Design! and is my attempt to make sense of what artificial intelligence — especially new tools like Claude Design — is doing to both design practice and the industry at large. The piece is pegged to the cheeky pop-up that Figma staged earlier in the year around the theme “design is dead.” That event, I thought, tapped into some deep-seeded thoughts I’ve seen in the design industry for as long as I’ve been practicing:

This feeling that design is dead, or at least dying, is a familiar one. I’ve worked across the design industry for nearly 20 years—from designer to writer to university professor—and for my entire career, death has seemingly been knocking at our door. 

I have a vivid memory of a high school friend (someone far outside my design world) asking me as we graduated whether I was worried about studying graphic design because anyone could get a copy of Photoshop. 

Then in the early 2010s, I remember conversations with designer friends about whether we should be worried about emerging platforms like Squarespace and Wix, which were making web design newly accessible to anyone with an internet connection, while a plethora of logo-making services were emerging, threatening the same for branding. 

Last month, when the man installing my new air conditioner asked what I did, and I told him I’m a designer, his first response was to ask how I was dealing with “this AI thing.” 

I’m sure we can go back even further. Before all of this, of course, there were designers who looked at the desktop computer with uncertainty. Some felt the same way about printing technology.

There’s a history of design that could be told through the lens of designers afraid that design is dying. Which is to say that this piece isn’t really about artificial intelligence or the impending death of design at all. It’s really about how design — in all forms, though I focus on my fields of graphic and interactive design — are always in a process of change, informed by technological innovation, cultural revolution, and more.

I was curious about how artificial intelligence was, perhaps, speeding up those fears so I spent the summer talking with the folks actually building these tools — the design lead at Anthropic, an insights manager at Figma, and a design VP at Amazon AWS — about how they use AI, how it’s changing their work, and what they see, from their roles, for the future of design.

I started working on this at the beginning of the summer as I was coming off another project that was informing much of my thinking. My day job is teaching the next generation of designers so part of my job also means not only that in some way, I have to believe design is not dead but that I also have to figure out what that future could look like.

One way I do that is running a sponsored studio project with SAS, an analytics software company based here in the Research Triangle, every Spring. I’ve led for the last five years and over the years, we’ve worked on projects around data ethics, new ways to visualize complexity, rethinking APIs, and new interaction models for generative AI. The outcome, because this is embedded in a design program, is usually framed as an interface: students research and think through big ideas and turn that into a coherent digital product. We write personas, develop user flows, create functioning prototypes, and film scenario videos. Having done the project for a few years now, both myself and my collaborators at SAS were interested in shaking up the project a bit. Some things we kept circling:

• how could we get students to think more broadly about their work beyond product-focused goals?
• were there other deliverables beyond the “interface” that could push students into areas of work and inquiry?
• what are new potentials for agentic AI and how might that inform what a designer works on?
• how do we build systemic thinking into the project from the start, considering various outcomes, stakeholders, and use cases?

The project we ran was called “Speculative Worlds: New Potentials in Democratic Practices”. Working in teams, students built fully-functioning AI agents using Github and VSCode and a combination of ChatGPT and Microsoft Co-Pilot to think through real, societal problems. They did research, built knowledge bases, iterated on policy design, and ultimately designed systems where people (citizens, public servants, etc) would interact with the agents.

The project shifted the focus of the students from immediately jumping to interfaces to spending more time on systems. For many of these students, too, it was their first time working — designing — in code. I think it’s fair to say it was a hard transition for the students: it they had to shift their mindset from “interface design” to “system design that happens to have some interfaces” (and those interfaces might take different forms: websites, apps, public signage, printed materials, etc.), they couldn’t jump right into Figma and start prototyping so the entire design process they were familiar with was flipped upside down.

I’ll have more to say and write about this project in the future but it clarified for me that the current design processes are too limited for the complexity of the problems, projects, and challenges designers are tasked with today. Code is, once again, a valuable design tool that students should have experience with. “User Experience”, the defacto design process and perspective of our time, already feels dated and quaint to me. One future of design, perhaps, is one moving upstream — away from the production of artifacts to the organization of systems. (This also speaks to another recurring research interest of mine: the rise of strategy as a separate field in design.)

I started writing this new essay with all this mind. I was thinking about the work my students did and how different it was from the work they’d done before but I was also thinking about the reception to that work in our final presentations which clarified new ways designers can and could work. I was also drawing on my previous Fast Company essays, on AI and aesthetics and on brand strategy.

This is, perhaps, a long-winded way of saying that I have no fear of AI killing design (spoiler alert!). To look back at history, is to see design — graphic, industrial, interactive — constantly adapting to increasingly democratizing tools. As I write:

In a paper presented at the very first Type Directors Club conference in 1958, British designer and writer Herbert Spencer said of the role of printing: “What I am trying to stress is that we must be constantly alert to the technical changes now taking place in printing and in methods of reproduction and mass communication generally.” Spencer’s urging feels just as relevant today, though the “technical changes” are quite different 70 years on.

The progression of technological advancements of the last, say, 30 years—from desktop publishing to Claude Design—has been a slow process of democratizing design tools by putting them in the hands of more and more people. The skills needed to produce design no longer require years of study as these tools become faster, cheaper, and easier to use.

What artificial intelligence does to design, then, is divorce conception from production (which, as I write elsewhere in the piece, is actually an old idea about design too!). It separates the process from the artifact. If an AI tool can make a decent layout, interface, logo, whatever that means more non-designers will start doing design BUT also that designers will start doing more “non-design” work. This is what my students did this past semester: research, policy writing, AI development, etc. AI doesn’t kill design but it does blur design and how designers work.

Alongside all this, I’ve continued my own mini-experiments in AI software design — building and coding little apps to fill gaps in my workflow — and it’s gotten me thinking about how, paradoxically, “user-centered design” is more important than ever before. AI is big: it thinks big, it likes big databases, it works well “at scale”. The designer, then, must bring this back to human-size. Design, then, perhaps can no longer be “universal” or “generic” or “standard”. Design can again be bespoke, personal, customizable.

This is what my students were doing in their project: they designed and built these big, complex systems but then they had to figure out how to make those legible for people. They had to figure out what interactions, tools, interfaces were needed for various people in various circumstances. In an era where it’s easier to produce design than ever before, the task before us is to figure out how to rise above the slop, to meet people where they are. As I close my essay:

This is the task before designers: How do we design large systems that work at a human scale, that truly put the human in human-centered design? That’s a task that’s as messy and unquantifiable as design itself. It’s a task that can’t be automated. It’s a task less about data and production and more about feeling. At the risk of correcting Jobs, design isn’t just what it looks like or how it works anymore: Design is how it feels.

You can read the whole piece over on Fast Company!