Should UX Designers use AI?
User experience is interesting. It does not focus on the tools, the techniques, it does not focus on the ego of the designer. User experience is pretty self explanatory.
This article is about my own experience and perspective on the hot subject of AI as a designer that started my career in 2012, when Smart Layers in Photoshop were the biggest thing in design tech.
Are you having trouble deciding on using AI or keeping it all manual?
I will use an example to break down this subject, hopefully in a simple way.
Imagine the best experience your website visitors or app users can have. Wow, right?
Ask yourself a few questions about it:
How did you get to that perfect experience?
The important thing here is to ask ourselves, who is asking that question – the designer/developer or the user? We must have a vision in mind on how to build that experience, and that vision is built by a constant back and forth communication with your user base. Be it through surveys, interviews, tests, what have you.
Do you have the time to build it and adjust it constantly?
Unless you have set aside time and space for the constant iterations needed, you might not have the free time needed for the manual constant iterations, ideation.
I myself designed thousands of screens, and all the time I was thinking – I wish I could get my vision out there faster. It is really not about me, nor the time spent in Figma, Photoshop, Affinity, whatever.
Do you want to use AI?
It is all personal preference of course. Imagine if you had an intern working for you, who is really knowledgeable about all the possible theories, the gestalt principles, the “psychology behind the decisions people make”.
Your own doctorate intern
The intern is a PHd in design theory, but it just lacks that certain “je ne sais quoi”, that you just can’t pin down without interfering and guiding him half the steps along the way.
Now, our AI tool intern builds at the speed of light.
It builds left and right, it builds wrong and it builds right, but the speed it can condense new information, present it, sometimes better than you, lay out the needed layouts, wow, imagine if you had that the whole time?
I know that a good UI designer can do a better job, but who is really stopping you?
If you don’t like what you see, do it better, or as a matter of fact, simply describe it better.
It listens, and it understands, and it understands exactly what you want.
Which AI tools should you use?
I personally use tools like Figma Make, Subframe.com, Claude.com, https://bolt.new powered by Claude , Weavy.ai which can generate and edit prototypes in minutes.
Also, from what I have seen on social media, there is a toolset that can make design systems and flows directly in your Figma Canvas. Keep in mind that this can also absolutely wreck your file if you are not careful, but luckily we have duplicate files or Figma File History too.
The cool tool in question:
Claude Code + Figma MCP Integration by Mark Cianfrani makes your figma file a playground for direct Claude generation
Figma also just introduced a way to turn the production code directly into Figma designs.
In short, it can do anything, and you can do anything you want to edit and adjust it.
You don’t like the buttons? Great, you can edit them through variables, each separately, or now you can just say exactly what kind of button you want, and the AI can do it for you.
The new unlimited world of interactions
Before, designers either had to code the interactions, design multiple component step figma animations, or have a deep knowledge of Adobe After effects to do any slider, button animation etc.
It was really time consuming to learn all that, and I know I spent months learning and using the tools on a surface level, unlike the professional animators, which I was not.
It gave me good results, but I had to spend half an hour or more on a simple slider animation.
Now, I can just say:
– Figma Make, can you do a slider animation + something previously impossible, like, have a glitch effect cover the previous banner text and turn it into the next one.
It will reluctantly but surely do so. The first results might not be perfect. But guess what, your users are now happier seeing what interesting glitch animation their gaming interface has, and you did it in a few minutes.
The results I got after waiting a few seconds are pretty astounding. If I wanted to do anything further, I would just press the “Copy Design” button, and carry on in the Figma Design mode.
I edited the suggestion it gave me to this, and am sending it back to Figma Make.
It Still kept the glitch animation intact.
Now if a client needed a gaming homepage slider example, it is pretty evident what it did for me in 3 minutes.
With my light interaction and my additional play on it, it looks totally different, and all while giving me the exact results I wanted.
How should you use AI tools?
It is just a matter of telling them what to do. It is all based on requirements, your user experience and interface design knowledge, giving them enough inspiration from a third source, or designing a few examples yourself.
As a UX Designer, I am not qualified to go in detail, but Claude Code, Bolt, V0 by Vercel, Lovable, all seem like perfect examples of what you could build with just a little coding knowledge to go alongside, as you would with Figma make and design knowledge.
