Using ChatGPT in User Experience Research — some resources
Powerful AI applications are being released and are reshaping the way we do business. I’ve been doing an analysis about how the UX field, particularly UX Research, could use these tools in our work and stay competitive. After playing around with ChatGPT, I did some reading to find out what other UX professionals are saying and what ideas and best practices have emerged so far. What I learned were a few principles to keep in mind when working with AI tools in general as well as a list of ideas for using ChatGPT (and other AI chat tools as they are released) in UX Research today. Of course, new tools are being released all the time, and our understanding changes as frequently as new insights are gathered and new capabilities are realized. It’s a very fluid landscape; this work is from a snapshot in early 2023.
Mental Models
New technologies require new mental models. “Killer machines coming for our jobs and lives” doesn’t seem very productive! I am more inclined to see my relationship with AI tools as one of collaboration. AI programs cannot do the job alone, as they require my ideas and judgment. But we can effort back and forth to create something faster, more accurate and creative together. I love this (short, video) description of that collaboration by Don Norman:
The role of Input — Output
Artificial Intelligence is created by humans and trained by inputting information from many, many sources — also created by humans. This means the data we put in and get out is only as good as we are. And we are full of biases. As in all our research endeavors, we must make sure we continue to design research tests with the highest standards — so the data we input to AI tools gives us good output. Diverse representation, accuracy and reducing bias are as important as ever to make sure we are not perpetuating systemic racism and marginalization of groups of people.
Ideas for using ChatGPT in user research
Here are 4 resources I found that highlight interesting and practical uses of ChatGPT in user research today.
When I started reading about ChatGPT and the content and deliverables it can generate, I heard phrases like ‘train the AI’ or ‘train ChatGPT on your dataset’ and was not entirely clear how to do that. So, I Googled it and found this informative video by Jason West. Not only does he give advice on how to train ChatGPT with your specific data, but also how to use a handy app to grab all the text from your website, making it easy to ‘feed’ to the AI.
This next 3 minute read by Harsh Solanki gives 3 ideas for using ChatGPT in your user research. They are focussed around the idea of using the AI to stand in for users. ChatGPT can be asked to simulate users by generating personas, responses to interview questions, and feedback on design concepts.
If there is a one stop shop for getting started using ChatGPT as a UX professional, this next find is it! Jeff Humble puts together an excellent resource including uses for both UX design and research. What are also super helpful in this article are the UX — specific prompt suggestions given to help get you started generating content.
Some of the suggested uses for UX research specifically are:
- Writing all kinds of content, including research questions or email introductions
- Searching
- Describing common knowledge
- Summarizing
- Analyzing and synthesizing datasets
Lastly, this 10 minute video from John Rodrigues is worth the watch. He has a ‘conversation’ with ChatGPT where he inputs his research and asks various questions, comparing his own insights with those the AI comes up with. He suggests and shows how to use ChatGPT to describe context and draw insights from user data, create user flows and quickly generate relevant copy to be used in place of lorem ipsum text.
I hope this article was useful for the UX professional beginning their journey with AI tools like ChatGPT. While not meant as an exhaustive list, the articles and videos referenced can be a starting point for framing our work with AI tools and making the most of the powerful assistance they offer.
— — —
I’m Kim Theisen, a freelance User Experience Researcher, writer, thinker, doer, speaker, and teacher. I have close to 20 years experience in helping people understand better ways to learn, conceptualize problems, and collaborate in equity. I love to build out practices and find that aha moment in every study I do. The opinions I put forth here on Medium are my own, and don’t represent any client or organization.
