Like much of the world, we have watched with anticipation and trepidation as AI has steadily gained in prevalence and intelligence. Is this the end of our industry as we know it?
I think not. At least not for now.
However, it is important that we learn to embrace this new technology. As is often said, it is not AI that will take our jobs but the people who are confident using it.
To us, the two key areas in which AI will enhance our work are efficiency improvement and enrichment of research insights.
Improvement to efficiency is the obvious application of AI. By saving time on repetitive, mundane tasks, we can free up the brainpower of our consultants to add the value where it’s really needed – delivering empathetic research moderation and conveying insight in a clear and actionable way to the people who need it.
The application of AI that is less obvious is how it can enrich our UX insights. Can AI extract more from qualitative research than a human being alone? Someone who holds all the research in their brain and is highly trained in its analysis? The experiments have begun here at WU-HQ, so watch this space!
How is AI helping UX today?
Unlike us, who are just starting out on this journey, many people are way ahead and already applying AI in interesting and innovative ways to enhance UX practices.
Qualifying quantitative research: a number of companies already offer enhanced surveying tools that offer us the chance to gain deeper qualitative insights, but on a quantitative scale. Tools like Inca and Bulbshare provide a chatbot style conversational interface that offers a much more engaging experience than your standard survey. Clever AI enhanced analytical tools in the backend also make it easier to find sense in all the data generated.
Accessibility testing: Automated accessibility tools are already being enhanced by AI to provide that subjective interpretation so often required of the Web Content Accessibility Guidelines. The accessiBe widget is just one example of this.
AI-led qualitative interviews: There are even some companies that offer AI-led qualitative research sessions, people like Bolt, who claim to be able to deliver test insights in 24 hours.
How can AI help UX in the future?
While these tools are exciting, they still feel limited. For AI to truly enhance the UX space it seems some serious progress is still needed.
Interpretation of behavioural insights: A key aspect of our work as human researchers is to both listen to what test participants say and watch what they do. Often, these two types of insight do not match up which is why it is so important to capture both. Currently, AI solutions seem to focus only on the attitudinal insights (what people say). Until AI can confidently interpret people’s behaviours as well (ideally without lots of fancy technology) we still have the upper hand.
Engage like a real human: For now, most AI conversations resemble a chat with an overly obliging customer services agent characterised by an eagerness to please that can often feel insincere. And as we saw recently in the news, AI is still only as good as the people who use and train it, which DPD found out to their detriment.
Protecting our testers and their data: The final, major concern is the level of confidence we have in the ethical and lawful management of our testers and their data by AI systems. Additionally, we often speak with vulnerable people or those that need added support due to access needs. I do not yet trust AI to treat these people with the care and respect they deserve and it’s likely there’s some way to go before I would.
For now, I feel confident that us humans have the upper hand on the machines. Our ability to read minute behavioural cues and adjust our research approach, ensuring comfort and confidence for all participants regardless of their background or needs, means it will be some time before we can hang up our UX hats.
Disclaimer: References to companies in this blog should not be considered an endorsement of their services.