Earlier this year, I had the privilege of attending and speaking at UX Y’all, a UX conference in Raleigh, NC. As you might expect, AI was a major theme and the focus of my talk. I ended up having many conversations with designers who had an interest in AI but shared questions about how to use it meaningfully in their work. I’ve felt the same way and am still working through those questions myself. In this post, I’ll share some common concerns I heard around AI and how I’ve approached those challenges.
AI is for prototyping, not production.
Many designers, myself included, see the value of using AI tools like Cursor and v0 to spin up clickable prototypes in minutes. What used to take days in Figma, InVision or Axure now take minutes. This has made AI especially helpful early in the design process during ideation and low-fidelity exploration.
A common question I heard was wondering how AI fits into later stages, like high-fidelity design. This is something I also face, relying on AI far less when creating polished mockups or production assets. What’s helped is focusing on where AI truly adds value, rather than forcing it into every step. Despite the marketing hype, AI can’t do everything (at least not yet), so leaning into its strengths, like rapid prototyping, goes a long way.
AI is too expensive.
Another common concern I heard was the cost of using multiple AI tools, which is all too valid. One speaker talked about AI being “in the hands of the people,” but an attendee rightly pointed out that this only feels true if you can actually afford the tools.
I’m fortunate that Atomic covers and encourages the use of AI at work, but even then we’re thoughtful about which tools we pay for. On the personal side, I’ve run into the same challenge of wanting to experiment without spending a lot.
I don’t have a perfect solution, but I’ve been able to do quite a lot with free tiers, especially with v0 and Cursor. They’re great for smaller design tasks. I only tend to hit limits when I try to build a full application, which is where I burn through credits. If I want to build a full application, I eventually need to pick a tool and pay for it.
AI struggles with legacy systems.
This is the concern I run into most often. AI is great for generating new concepts or fresh applications, but the truth is that many of us aren’t designing brand-new weather or habit-tracking apps. We’re joining existing teams and working within established codebases, design systems and products. Our ability to navigate these complex, legacy systems is something AI can’t match.
When working in these systems, I’ve had to figure out where AI genuinely helps and where the juice isn’t worth the squeeze. In her recent blog post, “Designers aren’t going away. Here’s why.,” Taylor captured this well:
“AI tools excel at the mechanical aspects of design, generating interfaces, variations, and code, but complex enterprise design has never been about production speed. It’s about navigating organizational complexity, understanding hidden contexts, and translating between conflicting stakeholder needs.”
AI can support our work, but it doesn’t need to be part of every step in the process.
A Thoughtful Path Forward
I’m grateful to the UX Y’all community for the opportunity to have these impactful discussion. It’s clear AI is becoming a powerful part of our design toolkit, but it isn’t a magic solution and it doesn’t need to be. The real opportunity lies in understanding where it genuinely adds value. As the tools evolve, so will our workflows, but our role as designers remains the same, bringing clarity and empathy to the products we create.