The Impact of AI on UX Design (Will AI Replace My Job?)
By Maren Lane
UX Y’all is an annual user experience conference that takes place in the Raleigh and Durham area, and it features UX speakers sharing their knowledge and experience in the industry. I attended the UX Y’all conference that was held in October 2023 at the Raleigh Union Station. The theme was “The Breaks” — stories of lucky breaks, putting on the brakes, things you have broken, and so forth.
One of the most intriguing topics presented at the conference was the impact of artificial intelligence on UX design and how it might “break” the field. The world has been abuzz with discussion about AI, its impressive capabilities, and how scared we should be for our jobs, so I was extremely interested to hear what industry experts had to say.
How Well Can AI Design?
Laura Ruel, a professor at UNC Chapel Hill, focused on comparing an AI-generated website design to a human-generated website design. The results of each were comparable in terms of visual appeal, navigation, interactivity, consistency, and credibility. However, Professor Ruel argued that although AI produces high quality design work, humans are still needed for soft skills such as communication, collaboration, empathy, and problem solving. These are the irreplaceable skills that designers should focus on developing, skills that are important for understanding people, strategizing, and working with other humans. She emphasized that it is unlikely that usability testing will be replaced by AI soon, as we will still be needed for understanding problems, user-centered design and identifying perception of inconsistencies. Only we can have an authentic understanding of the human problems of our users. And strategy, a highly complex human ability, is still something that AI cannot figure out, at least not yet.
Professor Ruel also talked about utilizing AI to generate a solid starting deliverable or launch point for a larger project. For example, a student studying climate change could use it to analyze data and generate insights on climate change, including links to relevant resources and even a research plan, all of which would be immensely helpful and timesaving. She said that in the future her students may be graded on how well they can write a prompt to feed to AI. As an example, a UX design student writing a prompt needs to include details like the type of product they are designing, whom they are designing for, and what use cases a feature will support. This detail and specificity determine the quality of the AI output. Learning how to skillfully utilize AI to brainstorm and create basic content will optimize our ability to work in the smartest and most efficient ways.
Where Do the Lines Blur?
University students are using AI and will continue to use it to help them in school, so the questions that students and users in general face will center around ethical issues in relation to crediting AI when it is used and determining what information from AI is accurate for use. The lines between what AI and humans create will become more blurred as we all increasingly interact with and use information that is AI-generated. It will be challenging but necessary to establish rules and limitations on the use of it in various environments and situations to protect the integrity, creative ownership, and trustworthiness of our work.
How Should We Use AI?
Dan Saffer, a professor at Carnegie Mellon University, echoed similar sentiments, cautioning against overestimating AI’s abilities. He began his presentation with some of the headlines floating around — “AI is overhyped” and “AI will take over our jobs.” He said that the greatest benefits from AI are not from all-knowing, all-powerful AI but from narrow AI that is focused on every day, mundane activities like typing an email, setting a reminder, creating a to-do list, or listening to music. This kind of AI is minimal risk, high value and does not require high technical performance.
Professor Saffer posited that companies are hunting in the wrong place for AI innovation; they are focusing on problems that are either high risk, technically challenging or require high accuracy. He gave the example of Spotify daylists, a new feature that provides users with curated playlists based on their listening patterns — low hanging fruit that is high value, low risk and requires moderate performance. The high-risk features like deepfakes or features that violate privacy are dangerous because their misuse can be very harmful. Saffer concluded by emphasizing that we should not hastily inflate AI because it fundamentally lacks common sense and context, so its use is unlikely to replace UX design jobs.
What’s Next for Designers?
Overall, I found the AI talks to be balanced in their presentation of the power and limitations of AI, as well as the fears and outcomes associated with the unknowns of this technology. While it can generate sophisticated design work and will continue to improve, human skills will never be replaced because human reasoning and creativity are hard to replicate and our soft skills in collaboration and communication are superior to technology. Instead of overestimating its capabilities, be aware that AI is primarily useful for augmenting and streamlining certain aspects of the UX design process like generating ideas and content rapidly. It is apparent that we will all need to utilize artificial intelligence to help us in our jobs as designers, but using it carefully and wisely is, I believe, where we will need to focus our attention to leverage AI as a useful tool that helps, but does not harm our work and the people we impact.
