Jun 17, 2026 · 23 min · 11 segments
Join Professor Judy Stephenson in conversation with Professor Dame Henrietta Moore, Professor of Culture, Philosophy and Design, and Director of the UCL Institute of Global Prosperity. In this…
Henrietta MooreGuest
Judy StephensonHost
to transform your job? Well, it depends what you mean by artificial intelligence.
So again, we do quite a lot of work in the IGP on artificial intelligence for social purpose.
But the reality is that ordinary language models are trained on what we've got out there.
And so if you're really dealing with a problem that you haven't seen before, AI is most unlikely to be very helpful with it, especially if it's a social problem.
So as we go on now to think about whether there will be something called general intelligence, then that's another matter.
But what most of us who use AI a bit, because we all have to, because we have to know what we're doing, and notice is that if you ask AI a very simple question from a research point of view, say, can you tell me about all the Gen Z protests that have taken place in Africa in the last five years? just provide the references, right? So on the first run, you will get most of what you might have got if you'd spent 25 minutes on the internet, and it does it in like four.
If you say, I don't really think you've got the right answer here, what I wanted was X, and ask another question, you might get a refinement, but you also might get a lie.
And you also might get data that's absolutely nothing to do with Africa, which is not a mistake that I would make if I was looking at it myself.
So by this time, I've asked it a question, it took four minutes and it took two minutes for the next one after that and three minutes.
So we're kind of heading up to 10 and I could probably do it myself in 22, but I wouldn't make any errors.
And then the thing about going to interview old ladies in Land Rovers in the middle of the night in Kenya, I don't think it does that that well.
to transform your job? Well, it depends what you mean by artificial intelligence.
So again, we do quite a lot of work in the IGP on artificial intelligence for social purpose.
But the reality is that ordinary language models are trained on what we've got out there.
And so if you're really dealing with a problem that you haven't seen before, AI is most unlikely to be very helpful with it, especially if it's a social problem.
So as we go on now to think about whether there will be something called general intelligence, then that's another matter.
But what most of us who use AI a bit, because we all have to, because we have to know what we're doing, and notice is that if you ask AI a very simple question from a research point of view, say, can you tell me about all the Gen Z protests that have taken place in Africa in the last five years? just provide the references, right? So on the first run, you will get most of what you might have got if you'd spent 25 minutes on the internet, and it does it in like four.
If you say, I don't really think you've got the right answer here, what I wanted was X, and ask another question, you might get a refinement, but you also might get a lie.
And you also might get data that's absolutely nothing to do with Africa, which is not a mistake that I would make if I was looking at it myself.
So by this time, I've asked it a question, it took four minutes and it took two minutes for the next one after that and three minutes.
So we're kind of heading up to 10 and I could probably do it myself in 22, but I wouldn't make any errors.
And then the thing about going to interview old ladies in Land Rovers in the middle of the night in Kenya, I don't think it does that that well.
The rest of this transcript — segmented and speaker-labeled, so you land on the exact moment something was said
Search every transcript — by keyword, by phrase, or by meaning, across every show Radar indexes
Trends — what is surging across podcasts, measured against its own baseline
Alerts — when a name you follow appears in a newly indexed episode
No account is needed to search Radar.
