Over the past couple of months I’ve had the opportunity to attend two very different conferences: UX Scotland and the Birmingham CST Data and Digital Transformation conference.
At UX Scotland, I was immersed in the world of designers, researchers, and practitioners embedded in UCD, with plenty of discussion on how AI is reshaping our profession. At CST in Birmingham, I was there with the DfE team, engaging with Trust leaders and executive teams (specifically those within Multi-Academy Trusts) and education professionals, talking through the opportunities and challenges of AI in education.
At first glance, these are very different worlds and use cases, but having worked across both industries, I felt connected to each conversation in a way I hadn’t expected.
What I was expecting was two very different sets of conversations, something that would challenge my thinking from opposite directions. What I got instead was one consistent message, arrived at independently by two very different rooms:
Start with the problem, don’t assume technology is the answer.
It seems to echo across both sectors. Don’t ask how we can use AI; ask what problem we’re trying to solve. The advice, in various forms, was consistent throughout:
- Technology should serve people, not the other way around.
- Define the outcome before you choose the tool.
- Sometimes AI isn’t the right answer. Sometimes it is.
It is easy to get caught up in the excitement of new technology, especially when it is moving as fast as AI is. It’s worth taking a breath before deciding on anything.
Plausible isn't the same as true
One of the talks at UX Scotland used a brilliant analogy from The Hitchhiker’s Guide to the Galaxy (Deep Thought analogy), Deep Thought is asked for the “answer to the ultimate question of Life, the Universe, and Everything.” After seven and a half million years of calculation, it produces the answer: 42.
The problem with this is that as Deep thought points out, the answer seems meaningless because those who asked the question never actually knew what the Ultimate Question was in the first place.
AI is very good at producing convincing-sounding answers. What it isn’t capable of, yet, is judging whether the original question was the right one, or whether the answer actually makes sense. That’s where human judgement still matters.
The speaker described AI as a brilliant but inexperienced intern. It can accelerate work dramatically, but you still wouldn’t send its first draft straight to a client. You, the human, remain accountable for what goes out the door and framing it that way makes the conversation feel more comfortable, a focus on the tool not a replacement for our judgement.
The skills we can’t afford to lose
At CST, Dr Mediha Khan compared AI to GPS. It didn’t stop us travelling when we adopted it, but it has slowly eroded our ability to navigate without it. A good visual analogy, and one that’s stuck with me.
AI has the potential to do the same to critical thinking if we don’t use it intentionally. Research shared at the conference pointed to students performing better with AI available, but struggling more once it’s taken away. The concern is not AI itself, but the over-reliance and the trust we place in it without question.
The challenge for education is making sure AI enhances capability rather than replacing it. As someone who has spent several years as an educator, I’ve seen how technology can narrow a student’s focus rather than expand it.
The accidental learning disappears once the tool hands you the direct route to the answer.
I sometimes think back to libraries where I seemed to spend half of my education. You never got straight to the thing you were looking for, and if you were anything like me, you’d get pulled into other ideas along the way. Turns out that detour was doing more work than it looked like.
User-centred design matters even more
One of my favourite sessions at UX Scotland was on leading a research community without formal authority, from Jessica Chapman.
The lesson wasn’t really about AI at all, but it landed there anyway:
Expertise isn’t having all the answers. It’s creating the conditions for other people to contribute.
It mirrors my own thinking about AI and design. The goal isn’t to replace human expertise, it’s to amplify it; the context, empathy, ethics and experience that let you recognise the unexpected insight that changes the direction of a service. (Turns out the library analogy meant something after all.)
Different sector, same principles
What surprised me most wasn’t how different the two conferences were, but how consistent the message was between them.
Designers and researchers talked about user-centred design, evidence and verification of ideas. Trust leaders talked about pedagogy and school improvements before technology. DfE talked about capability, safety and governance. Different terminology, same underlying concerns, an identical strategic challenge,
Whether you’re scaling a design methodology across a multidisciplinary organisation or integrating AI into your reporting across a network of academies, the task remains the same: embedding new technologies that align with your values while retaining consistency and delivering real impact to those who engage with your organisation.
As AI becomes more capable, value won’t come from simply knowing how to use the latest tool. It will come from asking better questions, exercising better judgement, and understanding where technology can genuinely improve outcomes for people.
That’s what makes AI exciting, and if anything, it’s what makes UCD more essential to us as designers, not less.
So what?
AI is already here, already changing how we work. There’s no denying that. But if I take one thing away from these two conferences, it’s this: slow down the instinct to reach for AI, and invest in the quality of your own thinking first.
- Start every AI idea with a problem statement. If you can't describe the problem in human terms, it isn't ready for AI.
- Treat AI outputs as a draft, not a decision. Build in the review, the challenge, the refinement, before it goes anywhere near a user or a student.
- Design for independence, not dependency. Always ask: what if AI disappeared tomorrow?
- Keep humans in the loop, especially wherever safety, fairness, learning or trust is on the line.
- Invest in thinking skills, not just technical skills. Critical thinking, questioning and domain expertise are worth more with AI in the room, not less.
- Design for the whole, not the individual. Avoid fragmented and individual silos by prioritising systemic governance, whether you’re deploying AI across a network of schools or a distributed design team.
We cannot resist AI; it is here. But we can be deliberate about how and when we use it. Those who are will be the ones who get the most out of it, in business and in education.
References
- Adams, D. (1979) The Hitchhiker’s Guide to the Galaxy. London: Pan Books.
- Boden, M.A. (2018) Artificial Intelligence: A Very Short Introduction. 2nd edn. Oxford: Oxford University Press.
- Kane, G.C., Palmer, D., Phillips, A.N., Kiron, D. and Buckley, N. (2019) The Technology Fallacy: How People Are the Real Key to Digital Transformation. Cambridge, MA: MIT Press.
- Khan, M. (n.d.) Research for the Good Future Foundation. (Accessed: 10 June 2026).
