One year of experimenting with AI, Figma Make, and trying to understand what still matters.
Last year, I traveled to San Francisco for Figma’s annual conference, Config. From the very first day, I felt the incredible power of what building in community can accomplish.
Sharing spaces of learning, reflection, and conversation with so many talented people was electrifying. I remember ending each day completely exhausted, yet filled with a renewed and almost overwhelming sense of passion. I met creatives from different organizations, founders, freelancers, and employees of large corporations, coming from different industries and different corners of the world.
When I returned home, I did so carrying countless ideas and projects I wanted to pursue, new tools I wanted to explore, and aspirations that made me love my profession even more.
I took something else back home, too: a growing anxiety about what was happening in the industry. The announcements made during Config felt like the starting point of what the following months would bring.
The AI Frenzy
Within just a few months, AI went from being embedded in our tools as a helpful assistant that handled boring tasks (renaming layers, generating images, filling prototypes with placeholder copy) to becoming the main actor. Suddenly, it wasn’t only helping us work. It was doing the parts many of us considered the most interesting and enjoyable. It was beginning to reshape certain roles and replace specific tasks across industries.
Figma Make, Figma’s solution for AI-powered creation, felt almost modest compared to what followed. Lovable, Claude, Cursor, Google Stitch. A growing collection of tools promising end-to-end solutions. Tools designed for one-person organizations, founders, startups, and anyone willing to experiment.
I felt compelled to create lists of tools and master all of them, because if you don’t know what’s out there, you don’t know what you’re missing. The problem is that today there are more tools being created than most people could ever realistically keep up with. As software creation becomes increasingly accessible, software itself is being produced at an exponential rate.
One Year of Experimenting
Over the past year, I tried to keep up.
I picked a handful of tools and focused on those. Figma and its ecosystem were obvious choices, but my main interest was always the journey from zero to one. I’ve always believed that an idea that never gets executed remains little more than a wish, and for the first time it felt possible to build and launch products with very limited resources.
The most difficult part wasn’t learning the tools, but deciding where to draw the line. AI is capable of generating ideas, interfaces, copy, code, and entire products. It can accelerate brainstorming, shorten execution, and remove much of the friction that once stood between an idea and a working prototype.
Over time I realized that the challenge wasn’t figuring out what AI could do, but deciding what I wanted to keep doing myself. Some of the most enjoyable parts of creative work happen before execution: thinking through a problem, exploring possibilities, making decisions, changing my mind, and sitting with uncertainty for a while before arriving at a solution. When I allowed AI to take over too much of that process, the results were often impressive, but they didn’t always feel like mine.
At the same time, when I intervened too little, I discovered something else: these tools are incredibly convincing. They generate outputs that look polished and complete, but that doesn’t necessarily mean they’re correct. The more capable they become, the more important judgment becomes.
After a year of experimenting, I found myself returning to the same conclusion over and over again. The problem comes first, as it always has. The tool comes second. And somewhere in between, there is still a place for human taste, intuition, and decision-making.
Config 2026, What Now?
This week, Config returns, and a lot has happened in just one year.
I believe Figma continues to bet on craft. It continues to bet on integrating AI into software that is ultimately operated by humans, on the value of collaboration, and on the strength of design communities. To me, that has always been part of Figma’s appeal and one of the reasons for its success.
I also believe there is still room for that kind of organization and that kind of product.
Not every company is a bootstrapped startup; not every team is a solo founder trying to do everything alone. Large organizations still exist, and they require tools that can support more complex workflows while remaining modern, integrated, and capable of reducing repetitive work and shortening iterative processes.
My expectation is that Figma will continue moving in that direction: embracing innovation while remaining within the ecosystem it has built.
I think it’s important to keep everything in perspective. As AI-generated products become increasingly mainstream and creation becomes more accessible, the challenge is no longer building something, but building something worth caring about.