“I must emphasize, no matter how obvious it sounds, that good curating depends upon a bottomless passion and curiosity for looking and questioning; and the desire to communicate that excitement.”
It’s been over a year since I have taken a moment to do a deeper reflection on current tech landscape, and I feel that it’s past due.
A lot has changed.
A lot of these changes can be attributed to the rise of generative AI, LLMs, agents, etc. But I feel it would be remiss to simplify the narrative to simply “AI changes EvRythinG!”
I think to understand the design and shipping of software in 2026, we need to understand history, systems, and frameworks and the problems we have sought to solve in the past few decades. And I feel that where we stand as a field is mostly due to automation sans curation.
When the Internet (with a capital I) started to gain mainstream attention (and before its bubble went pop), it was a bit of a Wild West. A lot of companies were excited about the implications of a new communication channel, but were still grasping as to how it can be leveraged to sell (sound familiar?). Building a site, let alone a .com business was fraught with challenges due to the tech’s newness. Building for non-standard browsers that each required a different approach. Needing to hand code practically everything. A lack of general agreement as to what made an experience “good.” One can easily think of this era and presume that it was fun, experimental, and better that the Internet we have now. But like everything, it was a bit of a mixed quality. What we lack in fun has been replaced with streamlined.
Agile. REACT/VUE/Angular. Design Systems. CSS/SASS. Figma.
Our tools and frameworks make life easier. Gen AI makes it exponentially easier. But sometimes the problem with easier is that a field can take an attribute like streamline, and take it wholesale. Agile becomes so fossilized that the incentive of good is replaced by the motivation to simply mark as done to start the next item. JS frameworks and design apps encourage us to care more to master the specific idiosyncrasies of these tools than to understand the deeper problems we solve. But with curation one can really, if one wants, hunker down and PLAY.
I recently was tasked with creating a prototype that worked as a stable client demo for a new feature. I confess that I am not a fan of Figma’s prototyping. Or many of the other prototyping tools I’ve used in the past. It’s not just clunky. It’s so limited that I end up distracted by the tool’s capabilities that I feel I lack freedom to explore the problem I need to solve.
I tried Figma Make to alleviate those pain points. I loved the experience. I felt the tool has a wonderful balance if one is willing to go to the file’s code tab. While Claude Design has a mess of nested divs (which might work for some but truly was a bit too much for me), seeing the style specs in each element made updates shockingly easy. I would prompt big changes, update the data arrays as needed. And in that I found the freedom to play. To really ask, what if we did this? Or what if we can experiment with this interaction. The scope of the small demo quickly expanded for me, and I was able to combine ole’ fashioned research, discovery via Claude, screen design in Figma, and fast iteration to design a new feature that made everyone excited.
I always loved analyzing the tightrope walk of automation and friction. And in that love I in time realized that there is no binary value to any one given tool. Be it a framework, design, app, or chatbot, a good interface does not magically spring from one of those tools. But if you treat the tools as a buffet to extract the most valuable attributes to bring into your process, you will benefit. Your work will be rich solutions to real issues delivered in the channel that best suits the task.