Most likely, over the past year, you’ve read hundreds of posts about integrating AI into your design process. It’s still relevant, but in my opinion, it’s the least exciting part for UX designers in the realm of emerging AI technology.
Share article
Most likely, over the past year, you’ve read hundreds of posts about integrating AI into your design process. It’s still relevant, but in my opinion, it’s the least exciting part for UX designers in the realm of emerging AI technology. Design experiments, new interface concepts in human-AI communication, collaboration with data science specialists, user research methods for AI, or AI design patterns are topics already visible on the horizon.
Last year, the world was taken by ChatGPT. Finally, each of us could interact with artificial intelligence. However, how do you evaluate the ChatGPT interface from a user interface design perspective? To me, personally, it resembles the time in history when computers emerged, and the first users interacted with them through a command line interface (CLI). Over time, creators realized that it wasn’t the most user-friendly way of communicating with machines, and graphical interfaces took over. At this moment, there is a race among technology companies to develop the best generative AI model (generating text, images, or videos), but there’s something more… Compare interfaces from OpenAI (ChatGPT, Dalle), Microsoft (Copilot, Microsoft Designer), or Google (Gemini, ImageFX), and see how they start experimenting with interface forms to improve communication between humans and AI. It’s a return to the essence of UX design and an incredibly exciting exercise in creativity. We’re at a stage where designers are once again brainstorming on a blank sheet of paper to develop breakthrough concepts.
Many designers conduct their own design experiments, which are increasingly breaking into the general discussion. Conceptual design, even without further interface implementation, allows a better understanding of the design requirements for AI and the user’s perspective when encountering completely new technology. Check out, for example, Maggie Appleton’s thoughts on AI Deamons, Branches, and Epi, as well as UX patterns for GenAI experience from Ryan Tang.
In the general hype around AI, more and more companies are considering whether their product can also benefit from this technology. Adding AI features can build a differentiator in user experience and elevate the product to a higher level, but can also ruin it, leaving a negative impression of the product innovation among users. You can check, for example, the recent widely echoed issues with historical character generation by Google’s Gemini.
Enabling UX designers to collaborate with specialists responsible for AI features in the product allows for a more holistic approach. The convergence of these two perspectives and close collaboration can support solving user problems in interaction with AI feature from both technical and UX design perspectives.
For instance the AI feature in product could be Miro Assist, which helps to better handle the analysis and synthesis of information. Another example could be the personalisation of learning processes in Duolingo.
Many problems have been difficult to solve so far, but thanks to the expansion of AI technology, suddenly it becomes possible. As a result, the number of startups pitching new ideas and building entire products around AI technology is growing rapidly. In this case, a product with an AI feature will still exist without AI. But a product built on AI without AI doesn’t make sense.
However, to create a holistic user experience while working with specialists responsible for AI technology, it will require designers to deepen their knowledge of various models so that they know what problems (technically) models can generate for users (behaviourally).
Ahead of us are also the evolution of user research methods, which will provide valuable observations and feedback to the data science team, the development of workflow with AI specialists, and the discovery of new design patterns. In my opinion, this is one of the most exciting stages on the horizon.
Examples of products built around AI technology include Atomic AI, which may herald a revolution in medicine, or AlphaSense, which changes the traditional approach to market research.
Incorporating AI tools into the design flow can significantly speed up your design work. But… critical thinking is an integral part of this process. Generating personas or sample user responses won’t get you closer to solving the problem and may only deepen your bias. Instead, testing different AI solutions in various situations will help you develop your own AI workflow. The Interaction Design Foundation’s platform offers a great AI for Designers course that walks you through the various tools that support the design process, but also highlights the various pitfalls you can fall into if you don’t approach them critically.Examples of tools you can use in your daily work include Uizard or Galileo AI. They will help you in the ideation process and finding UI solutions.
