There’s a recurring take doing the rounds on socials. Thought leaders declaring that AI has killed the design process (e.g. Lenny’s podcast).
I personally think they’re wrong. But more importantly, I think most people (including designers) don’t know what they mean when they say “design process.”
When pushed for specifics what they’re actually describing is making static frames in {insert your design tool of choice here}. Creating screens, components, spaghetti prototype flows and packaging them neatly to throw to an engineer.
That was never the design process.
What the Design Process Actually Is
The design process has always been about reducing uncertainty. Typically addressing variations of:
- Are we building the right thing? (Problem space)
- Are we building it right? (Solution space)
It’s worth noting that software didn’t invent design. Humans have been building things with economic intent for millennia. Every one of them passed through some form design process before reaching final form.
From a modern, software lens, design artefacts surrounding the final feature or product e.g. interview transcripts, wireframes, Figma prototypes was (or at least should have been) a means to answering one of the two critical design questions.
Figma did an exceptional job of marketing themselves as where design happens. “Design” collapsed into “what you do in Figma.” They attempted turned a tool into a category and a generation of designers grew up inside that frame.
I’ve written about this pattern before in Tools ≠ Skills. When a company successfully convinces people that mastering their tool is itself a valuable skill, those people become advocates for the tool’s continued relevance. It’s rational self-interest, but it distorts and dilutes the field of design.
The worrying part isn’t that junior designers think this way. It’s that senior ones do too.
The Impact of AI
AI is not disrupting the design process. It’s giving us the chance to execute it better than we ever could before.
The design process has always demanded that we get ideas in front of real users as quickly as possible. In the past, we were limited by both time and cost constraints. Lower fidelity, higher abstraction, more leaps of understanding were required from the user you’re testing with. The feedback was always noisier than we’d like.
Now the feedback loop is tighter. We can get to higher fidelity faster, which means we can validate earlier and with less ambiguity. We can test whether we’re building something in the right way with multiple prototypes that go beyond click-only prototypes and include the data that underlies the application.
The design process is now empowering designers to take their proposed solutions and validate directly them without being bottlenecked by engineering teams or unsatisfactory abstractions.
The cost of building has been reduced >10x so its more critical than ever to build the right things in order to being competitive.
Same Process, Different Skills?
If AI compresses time-to-prototype, the process doesn’t change. But the required skill distribution and willingness to adapt inside a design team might.
There’s a difference between someone whose primary value-add was Figma mock-up production and someone whose value was always the quality of their thinking about what to build and how to build it. It’s worth being precise about which conversation we’re having.
The first group is and should be disrupted. The second group just got a significant upgrade in their process.
The design process isn’t dead. It just has better tools.