AI Can Do My Job Now (Sort Of) (Not Really) (It's Complicated): A user researcher's complicated relationship with AI
Time: Wednesday 13th May 14:30-15:00
Room: Loft
Track: User Experience & Content
I've been doing user research for about 15 years now, across government, health, disability, education, all sorts of domains. For most of that time, research has been a discipline that was expensive, time-consuming, and gated behind a lot of process and methodological rigour, which was there for good reasons but also meant that a lot of teams just couldn't do it, or couldn't do enough of it.
AI is changing that pretty fast. I've been using AI tools in my research practice over the past couple of years and some of it has been really useful. Things like synthesis, first-pass analysis, and drafting that used to take days can happen in hours now. Research that would have been too expensive for certain projects is suddenly possible, and I think that's worth being excited about.
But I've also seen AI get things confidently wrong in ways that are hard to spot if you don't already know what good research looks like, and I think that's the tension worth talking about. If AI makes research accessible to teams who couldn't afford it before, but they don't have the experience to know when the output is rubbish, are we actually better off?
I'll share the tools I've been using and where I think they're great, the things I'm still worried about (especially around false confidence and the appearance of rigour), and the parts of research I'm not willing to hand over, like the actual conversations with people, the judgment calls about what matters versus what's noise, and the "so what" of turning findings into something useful. This is a bit of a commentary on where the research discipline is heading and what we need to hold onto as things get faster and cheaper.
What people will take away:
- AI tools and approaches that are actually working well for research right now
- Where AI falls over and how to spot it
- What parts of research still need a human, and why
- Some questions about the future of the discipline that are worth sitting with
AI is changing that pretty fast. I've been using AI tools in my research practice over the past couple of years and some of it has been really useful. Things like synthesis, first-pass analysis, and drafting that used to take days can happen in hours now. Research that would have been too expensive for certain projects is suddenly possible, and I think that's worth being excited about.
But I've also seen AI get things confidently wrong in ways that are hard to spot if you don't already know what good research looks like, and I think that's the tension worth talking about. If AI makes research accessible to teams who couldn't afford it before, but they don't have the experience to know when the output is rubbish, are we actually better off?
I'll share the tools I've been using and where I think they're great, the things I'm still worried about (especially around false confidence and the appearance of rigour), and the parts of research I'm not willing to hand over, like the actual conversations with people, the judgment calls about what matters versus what's noise, and the "so what" of turning findings into something useful. This is a bit of a commentary on where the research discipline is heading and what we need to hold onto as things get faster and cheaper.
What people will take away:
- AI tools and approaches that are actually working well for research right now
- Where AI falls over and how to spot it
- What parts of research still need a human, and why
- Some questions about the future of the discipline that are worth sitting with
Speakers
Billie-Mae Kennedy
Mae is a human-centred-designer who cares making government services better for the people who depend on them. She's worked across both the public service and consulting on some of the most complex government services in Australia: aged care, disability, gender-based violence, health, education, cybersecurity, water and clean energy.
Mae has led design practices and multidisciplinary teams across service design, product, and policy, including a fair few websites. She's particularly interested in the ethics of how we design and build things, building vibey teams, and embedding continuous discovery into everything.
Mae has led design practices and multidisciplinary teams across service design, product, and policy, including a fair few websites. She's particularly interested in the ethics of how we design and build things, building vibey teams, and embedding continuous discovery into everything.