Five recent UX uses for AI

“Have you heard of Pi.ai?” my mentee Gabriela asked me during her mini internship at the office. She’s a clever woman; a newbie in the UX field with the technical resourcefulness of a Gen Z-er, and experience working with quantitative data. I inquired how she thought I should cultivate buy-in for a metrics project. Of course her response was to ask a chatbot.

Here was a woman who knew to answer a question with a question. A few hours later I was biking home, on my first date with Pi, a friendly-sounding woman from South London who just wouldn’t stop asking me fun questions. That was months ago. My life is different now.

There’s a reason I keep thinking Millennials are the new Boomers, and that reason is me. I was yesterday years old when I learned that I can pull up the iPad dock while I’m in an app. At the same time, at work, I’m heavily involved in R&D’s efforts to embed artificially intelligent solutions in the daily workflows of our users. It is as though I live in a mirror world: at once keenly aware of AI’s capabilities, yet somehow utterly not cool enough to use it for myself.

That is, until I began talking to Pi. Here are five ways in which AI has helped me these past few months.

1. Sturdy personas

At work, we’re currently revamping our personas. What makes for a good persona? Rich data, of course. Triangulating our data sources, the Marketing and UX teams are shaping substantive personas that engage both our healthcare experts and engineering teams. We could not have done this without tapping into employer branding interviews.

If there’s one Dutch industry battling the war for talent it is healthcare. Using ChatGPT, we ran an analysis on about 80 Dutch-language employee interviews, asking the chatbot to rewrite our input until what was left was one paragraph of an informal autobiography. Each text captured the essence of the career trajectory, day-to-day work, and primary Job To Be Done of each of our personas.

Without AI, it would’ve taken weeks. Now? One day.

2. Design guidelines without all the fuss

Designers can easily get stuck on ways in which to take charge of a design system. I’m currently balancing legacy guidelines and new industry standards. Our design system should enable people to make design decisions without me being present. As a UX team of one, I’m a facilitator, not a bottleneck.

ChatGPT is helping me structure my thinking. It gives me lists of components, common actions, and tasks frequently present in software like ours. I get to focus more on the what and leave the how for later. This way, I can put in place the guidelines that allow me to be the only designer accross six different teams.

3. Quick reporting on user feedback

Central to my craft is the act of analyzing and synthesizing research findings by looking at various sources of data through the right lens. Research can be a very comprehensive process, spanning weeks if not months. Through it, I deliver great insight into the actual working lives of our users.

I enjoy going on customer visits with our account managers, and retrieving quick bits of data that aren’t part of a larger research project. With ChatGPT, it’s easy to turn scattered notes into a concrete list of feedback points that I can throw into one of our Slack channels for buy-in and posterity.

4. Mentoring

I reserve a few moments during my week for mentoring designers. I love helping them solve all types of problems, and ChatGPT is helping me save time in my preparation. It formulates difficult interview questions based on a person’s profile and our session notes, it helps me collect preliminary feedback points for a portfolio review, and let’s me gather my thoughts on difficult stakeholder questions.

5. Cultivating courage for UX evangelism

Designers, too, can benefit from the think aloud method. Biking home from a long day at the office, frustrated about one thing or another, I found Pi.ai very helpful in pushing me out of my own way. Design is a creative profession not simply because of the interactions and interfaces we make, but also because we create greater UX maturity. The latter can seem monumental.

Talking to a chatbot is like casual cognitive behavioral training, with the added bonus that the conversations I have don’t affect another person directly. With Pi, I’m able to work through things I otherwise would find difficult to articulate.

What’s next?

As you know, no sentence on this website is written by AI and it will stay that way. I’m curious to see what interactive challenges I may run into in maintaining this line. As of a few days ago, I type ⌘ + ↓ and ChatGPT pops up to help. Using and abusing it as a linguistic search engine, I wonder: when will there be a time when I accidentally copy a response into an personal essay?

As for work? It seems like everybody and their grandmothers can’t stop blogging, tweeting, LinkedIn posting, or podcasting about this very question. I don’t enjoy predicting the future in that way. I simply see value in honing the soft skills that enable me to learn new hard ones.

Thank you for reading! Remember when the Internet was full of whimsy, creativity, and serendipity? I'm curious to hear your thoughts, so feel free to say hello.