If you're a designer right now, some of these probably sound familiar:
- There are just too many tools to choose from, and they're constantly changing.
- I don't know when or how to use them in my workflow.
- I don't have the technical knowledge to use them properly.
- Am I falling behind?
You're not alone. UX Tools 2026 survey found that 37.7% of designers do 0 vibe coding, and 30.5% want to build with AI but haven't started yet. The pressure is real, and the noise around which tools we should be using changes every day.
The core idea: be intentional
The goal isn't to use more AI. The goal is to be intentional about how AI shows up in your work.
There's a useful distinction between speed that is helpful and speed for speed's sake. Going faster only matters if you're going somewhere worth going. Otherwise you're just generating more, faster, with less thought.
Three questions cut through most of the noise:
- What is AI actually good at?
- Where does it fall short?
- How might it fit into my workflow?
What AI is good at
Moving fast. AI gets you from an idea to something you can quickly try out. Iterate quickly, explore through lots of ideas. David, a Lead Product Designer at CallRail, uses it to move quickly from concept to implementation and cut down manual work.
Surfacing context. Carlo, a Principal Product Designer at Lady Bandit Studios, uses AI to analyze and search across piles of docs, extract the key info, and produce summary artifacts. Hours of reading turn into minutes.
Unlocking new capabilities for designers. Things that used to require a different skill set are now in reach. It can teach you while you work, help you contribute to code even if you're not a developer (depends on your dev 😅), or carry a project from concept to working product end-to-end. We’re no longer bound to just the idea and blueprint side of building. We can go in and contribute to the actual building of our designs.
Quick prototyping. A working prototype used to be a real project. It could take days of coordination and technical effort. Now it can fit into an afternoon. Mizen was mostly designed and built by me and a few others. There’s less and less in between ideas and getting it in front of people
What designers bring
This is the part that gets lost in hype. AI doesn't replace what a designer actually does.
Building foundation. Sean Loose put it well: "You don't open a design tool first. You sketch, you draw, you build things, you look at what others have made. The tools you reach for shape the way you think, and skipping this part is how you end up with design that's unintentional."
True empathy for the people you're making things for.
Human context. Christina, a UX Designer at eWorld, points out that projects have existing context AI doesn't have, and AI-generated mockups don't fully align with the level of system thinking real product work requires.
Sweating the details. The last 5% that makes something feel finished. AI can’t sweat this on its own.
Picking tools without losing your mind
Tools get easier to navigate when you put them in categories. The way I think about it, there are roughly three buckets:
| Fidelity | Tools | Advantage | When to use it |
|---|---|---|---|
| Low fi | Figma, design canvases | Fast, simple, focused | Early ideas |
| Mid fi | Lovable, Figma Make, v0 | Feel how it works, easier to share | Exploration and testing |
| High fi | Cursor, Claude Code, VS Code | See it in real context | Validating |
If you're brainstorming, reach for a canvas tool. If you have some technical knowledge and want working code, reach for a code editor. The decision gets a lot easier when you stop comparing tools across categories and start asking, "What kind of work am I doing right now?"
BYOT: bring your own tools
Your toolkit doesn't need to look like anyone else's. The categories matter; the specific tools inside them are personal.
Pick one (or two) from each bucket and actually learn them:
- A design canvas tool for thinking and shaping ideas
- A no-code / sandbox tool for getting something interactive in your hands
- A code editor tool for working in real context (I run localhost in Conductor and preview changes in the browser — small loop, fast feedback)
Mines is Paper → v0 → Codex. Yours might be Figma → Lovable → Cursor, or something else. The point isn’t to match someone else’s stack. It’s to have a stack you’ve built real habits around.
Four tips for choosing tools
- Balance learning with mastering. Learn new tools, but actually get good at the ones you keep. There's a difference between collecting tools and having a toolkit.
- Learn from others. People around you are figuring this out in public. Watch how they use things. Steal what works. Skip what doesn't.
- Develop your taste. Your trained sense of what's good and what's bad. You build it by surrounding yourself with great work and great community, and by thinking critically about what you see. Then experiment, and use that experience to form your own opinions.
When you're evaluating a new tool, Claire Nicholson has a useful frame: define your standards (accuracy, bias, security, whatever matters for the work) and assess the evidence.
When Michelle was testing Linear, our standards were: a tool built specifically for product work, integrated with our other tools and how we like to build (Notion, Granola, GitHub), and relatively low cost (the free plan works for our use cases for now). A quick trial run, using it with my side projects, proved to us that this would mean spending less time issue tracking and more time creating!
Floor vs. ceiling
All of this, the categories, the taste, the intentionality, matters because AI is pulling the work in two directions at once.
- The floor: More people can produce something passable.
- The ceiling: People who know what they're doing can go further than before.
You get to choose which one you're optimizing for.
Closing
The design process is shifting. New tools, new ways teams collaborate, new things that are suddenly possible. But the essence of design hasn't changed: care deeply, understand the context, and make great software for everyone.
The goal isn't to use AI. It’s to do more quality work.
Sources and inspiration that shaped my thinking:
- Writing & thinking — Proof of Concept / David Hoang
- Design engineering — Dive Club, Julius Tarng, Cursor Learn, the d×e course
- Craft & polish — Emil Kowalski, Josh Puckett, Devouring Details
- Culture — Family Values, paula on X
My current toolkit lives at gageminamoto.com/tools.