Something’s been stuck in my head since attending Config: the crowd’s outbreak of applause watching AI “Rename Layers” vs anxious hush watching AI “Make Designs”.
In general, I’m an AI-optimist with design. I think there’s a ton of room for it to make our lives easier: freeing up our time, helping us rapidly iterate, and providing heuristic-based feedback. Those actions align well with my goals leading DesignOps: enable my team to focus time on their craft and do their best work.
That said, there’s an unspoken anxiety watching AI become increasingly aimed at driving craft. Watching an AI crank out a hi-fi UI in 10 seconds invokes questions on job security, commoditizing design work, and bypassing critical problem-solving in favor of polished visuals. Anecdotally, I’ve observed designers have different comfort zones with how involved AI is in their craft. The more tedious or mundane the work is - sure, let AI automate it all intelligently. Hence Figma’s “Rename Layers” and “Visual Search” being so well received, as this work is not part of the craft itself.
As we start to move away from tedium and towards craft, AI can comfortably take on the role of an assistant, enabling designers to iterate quicker and keep them in the flow. These are things in the canvas - generating intelligent placeholder text, suggesting alternative interaction design patterns, helping us build components or layouts, and so on. I think we've only scraped the surface of how AI can contribute value here. At the pure craft level, I personally believe designers will (for now!) be most receptive to AI as an advisor. Let me run an AI-driven heuristic evaluation, surface accessibility issues in real time, critique my work, etc. When AI takes on a bigger role here, we quickly enter an anxiety zone. Figma spoke to “Make Designs” helping designers get past the blank canvas, which is valid but the reality of Product Design is you don’t go straight from blank canvas to hi-fi mocks; in doing so we perpetuate that fallacy to non-design stakeholders. “Can you just mock this up?” now has a clear AI answer. I’d have liked “Make Designs” to first focus on generating wireframes or 'translating' them to hi-fi with a design system. Or perhaps riffing multiple alternatives to what a designer is currently exploring. IMO, I think both of these use cases would have been more readily embraced by the community at this point in time. I’m sure these comfort zones will blur and evolve, so I'm eager to see how Figma adapts based on customer feedback!
More Thoughts
