30 ways to use AI in UX research and design
An UPDATED brief summary
Why this article is important?
UXers must have a say when it comes to creating AI experiences that are fair and accessible to users.
In 2023, Jakob Nielsen in the article UX needs a sense of urgency about AI, wrote: “As UXers busy themselves elsewhere, engineers are shaping the future: AI-driven user experiences. I firmly believe that AI holds the key to the future of UX. UX needs an espresso shot of urgency around AI for two reasons: the quality of user experiences globally and your career advancement. Let's dissect each of these”.
Since reading it, a little voice in my head has been telling me I should write about AI in UX to help the community thrive. The article was so impactful that I created this substack.
So, in this article, I will share 30 ways to use AI in UX.
This list is the result of my experience as a UXer, embracing all these new possibilities of using AI at work.
I hope it will help you!
The 30 ways to use AI in UX workflows
Detect bias in AI
In UX design — Google Fairness Indicator evaluates machine learning algorithms that are used in UX design that customize content on websites according to user's preferences, such as recommendations that are biased towards a specific user profile.
In copyright and content — MonkeyLearn and Acrolinx: detects biases in UX copyright, design, and images.
In machine learning algorithms — the IBM 360 toolkit is a free toolkit that helps create more transparent machine learning algorithms.
Creative inspiration for creating images and videos
I believe these generative AI tools are great for creative inspiration when it comes to:
Creating images — Leonardo.ai, Openart, and Picsart will help you generate images.
Generating videos with text — with the use of AI, Pictoryai, fliki will help create a video using text prompts.
Forecasting future scenarios
Pre-mortem analysis — when creating a new product/design, you can use AI to forecast potential risks in advance. An example of a prompt that you can use with juliusai, ChatGPT, and Copilot: "I am working on a digital product [give context about the project: who uses it; what it is: app, site, kiosk; where: if applicable the field it operates; when was created; and what's the motivation/problem you are trying to solve with it]. Can you run a premortem analysis?"
Identify specific requirements for developing a product/design according to a framework or best practices — Utilise juliusai, ChatGPT, and Copilot with the following prompt:"I am working on a digital product [give context about the project: who uses it; what it is: app, site, kiosk; where: if applicable the field it operates; when was created; and what's the motivation/problem you are trying to solve with it]. Can you suggest software and hardware requirements I need to have in mind according to the 10 usability heuristics by Jakob Nielsen?"
The same tools can also be used for a Pros and cons analysis — an illustrative prompt that you can use would be: "I am working on a digital product [give context about the project: who uses it; what it is: app, site, kiosk; where: if applicable the field it operates; when was created; and what's the motivation/problem you are trying to solve with it]. What are the pros and cons?
Analyzing qualitative information
Thematic analysis of information — juliusai, ChatGPT, Copilot can be instrumental in conducting a qualitative analysis according to specific themes. Make sure you read my article on How to use Chatbots to fastenQualitative data analysis to avoid biased analysis and results.
Understand what users are saying about a product.
AI can help you understand what users are saying about your product — tools like Kraftul analyze users' feedback from multiple sources, like: Slack, Reddit, Google Play, App Store, and many more.
Improve communication and tone
Improve communication — ask yourself what you want to communicate and then ask juliusai, ChatGPT, Copilotto help you be clearer with a specific stakeholder profile.
Adjust writing/speaking tone — if you want to speak more formally or casually, you can use juliusai, ChatGPT, and Copilot to adjust the text to a specific communication tone.
Speak or write in a different language — the same conversational chatbots mentioned above can help you translate text, or write in different languages.
Improving your grammar in different languages — tools like Grammarly use AI to check your grammar and work as your writing assistant.
Create video presentations
Translate text to video — Synthesia.io uses AI to help you transform text into a video in minutes.
Collecting and synthesizing academic research
Collecting academic sources of information — ChatGPT created a new feature (chatGPT academics) that allows you to combine academic sources of information.
Synthesize the sources of information — after collecting these sources of information, you can easily ask chatGPT to synthesize information in a specific format.
For creating UX deliverables:
Mindmapping — Figma will help you create mind mapping with the help of AI. Use the mindmap map feature when you need new ideas about a specific subject.
For creating user journeys — with the "update from your computer", update user interviews, or past research about a specific study into the chatGPT, then you can ask the chatbot to create a journey map.
For reflection by user impersonation — ask Copilot, ChatGPT, or other conversational chatbot to impersonate the role of a UX researcher (or any other useful role: product manager, UX designer) when creating a specific deliverable, this will help you reflect and gather ideas when creating your version of the same deliverable.
Generate UX & UI designs.
Generate design templates FigJam AI — use Figma’s AI tool to generate design templates, organize and brainstorm ideas, and automate repetitive tasks like preparing boards.
AI tools for customizing the user experience delivered to the users
Adapting content based on user interactions — Pendo AI, Algolia, Dynamic Yield automatically rearrange content in designs: grid, menu items, and buttons to deliver customized layouts according to each user preference.
Optimizing UX and product copy
Ensure copy consistency across all UX designs — Contexta, Unbounce Smart Copy, Copy.ai use generative AI to give text suggestions based on best practises, demographics, and user preferences.
AI tools for UX design prototyping and testing
AI to transform sketches into functional prototypes — Uizard, Fronty both use machine learning according to design trends and user patterns to generate HTML and CSS.
Conduct UX research
Moderation with a non-human entity — this softwares use AI (specifically a non-human moderator, e.g. avatar in place of a human moderator) to conduct interviews, surveys, and co-creation sessions. Feel free to explore tools like: askmore, boltchatai, voicepainel.
Synthetic user research — tools like: Neurons, Attention Insight, Synthetic mimic user behavior and predict their behaviour. These tools allow testing with artificial/synthetic users based on AI-automated models. It seems helpful for testing preliminary concepts (for example: identifying specific critical user scenarios) and then expanding and testing with real users.
Biometrics tools — Tobii, iMotions, RealEye, Noldus, loceye analyze users' facial expressions, or voice patterns to create heatmaps or visualizations of subconscious behavior in response to a given design.
Delegate simpler tasks to your “AI trainee”
AI research collaborator — tools like ChatGPT and copilot are here to be our collaborators. These AI companions will help researchers and designers analyze information in large volumes, summarize data, or get inspiration for creating a new design.
Improve accessibility