Welcome back to another edition of C&F Talks. It is a great pleasure and an honour today to be interviewing Professor Sir Peter Mathieson, who's a Principal and Vice-Chancellor at the University of Edinburgh. Peter's going to be speaking at the National AI Policy, Infrastructure and Skills Summit being held in London on the 12th of November. Peter, welcome.

Thank you very much, Maurice. Great pleasure to have the chance to talk to you.

Fantastic to have you with us.

Please, could you give us some background first on both the University of Edinburgh's EPCC, the Edinburgh Parallel Computing Centre, which is the UK's first national supercomputing centre, and more generally, on Edinburgh's world-leading capabilities in supercomputing and AI.

Yeah, thank you very much. I mean, the University of Edinburgh is very proud of its history in artificial intelligence. So, we were the second university on the planet to teach AI.

The first was Stanford back in the 1960s, and Edinburgh, in collaboration with Stanford, right from that time, has been actively involved in the teaching and the research of artificial intelligence. Last year, we celebrated 60 years of teaching AI. So, we were teaching it right through the time when AI was unfashionable, and obviously now everybody's now waking up to the issues, both the risks and the opportunities around AI, and Edinburgh's very proud to be part of that kind of thought leadership.

We have the largest establishment of computer science researchers of any university in Europe, and we were ranked number one in computer science in the last two research excellence framework exercises. So, we have some credentials in the past. What we've tried to do now and in the future is think about the gaps.

Where are the gaps in AI? We think that one of the big gaps is around the ethics and the regulation and the law of AI, and we invested in that about six or seven years ago, actually before ChatGPT arrived, and that now looks like a very good decision. So, we have a lot of expertise, I think, in some of the technology, but also in some of the sort of broader societal implications.

When it comes to the supercomputer, so Edinburgh has been the home of the UK's national supercomputer for about 25 years. We had Archer, and we have its successor, Archer 2, which is the current national supercomputer, and just recently, the UK government has agreed to fund the next generation of supercomputer here in Edinburgh, and we are very honoured by that trust that the government has shown us. We believe that we have the right staff and the right experience and the right kind of infrastructure to provide a national resource in supercomputing, which happens to be based in Edinburgh, but it's obviously not just for Edinburgh, it's for the whole of the UK.

And that's a very impressive track record, going back 60 years, as you say, and very impressive in terms of the range of capabilities you have in this area.

The EPCC recently won EU and UK government funding to establish and operate the 10-million-euro UK AI factory antennae. What is the UKAIFA, and how would it boost adoption of AI among businesses and researchers across the UK to help unlock untapped economic and scientific potential?

So, this is a collaboration with the top centre in Germany, so we're effectively being designated as the two centres in Europe to lead the development of AI, the understanding of how it can be applied to society's great challenges, and also to understand what the risks are and how can we mitigate them. So, I think this is, it's separate to the supercomputer decision in many ways. I mean, there's obviously an overlap between supercomputing and AI, and I should say, Maurice, I am not an expert on AI, I'm a kidney doctor trying to run a university, but I've learned a lot about AI capability in the last few years that I've been here.

And my colleague, Mark Parsons, who's the Head of EPCC, makes it very clear that supercomputing and AI are different but complementary. So, you need to have supercomputing capability in order to be able to properly train and interrogate AI applications. And so, these two things go together.

And I think the partnership with the German group is, it already exists, we're already working with them, but I think this additional funding and this designation that we've just received will enhance our capability to bring the benefits of both AI and supercomputing to a much broader set of clients, as it were, or set of users. I think it's about training, it's about understanding what the boundaries are, what still needs developing further. And I think the putting together of great minds in two of Europe's great centres has to be greater than the sum of its parts.

No, absolutely. And it's good to see an EU-UK collaboration in this space, I think, given the years-

Yeah, I mean, I think the, in fact, I've just come back from Brussels where I was leading a University Scotland delegation. And although the UK is no longer a member of the EU, I think we are treated with respect by EU partners. I think the UK is still seen as a source of great expertise and a very sort of desired partner in many ways.

And I think this latest funding announcement, which is, as you say, only just been made this week, is another manifestation of that. So, despite the politics of the situation and us no longer being in the EU, where there's genuine complementarity and genuine need for the expertise that exists in the UK, there are still mechanisms whereby that can be achieved.

Yeah, no, I think it's fantastic to see that sort of pan-European, if you like, collaboration, including the UK.

The significant investment in the UKAIFA shows the important role that universities have in deepening our understanding of cutting-edge technologies and placing a trade, if you like, for industry and the public sector, which will hopefully lead to economic growth and job creation. How do you see the relationship between universities, government, and the private sector developing in terms of fostering the take-up of growth of AI in the UK?

Well, I think it's essential. I do believe that universities bring something very significant to that relationship. I mean, there is a bit of a tendency to think of universities as just being big schools.

So, we teach larger classes, and we teach slightly older people than schools. But I see universities do much more than that. We are obviously engines of research and a lot of interest in the commercialization of our research and the creation of new companies, the development of patents or intellectual property, et cetera.

But we also want to collaborate with industry, and we want to collaborate with government. I'll say a word about government in a second, but in terms of the collaboration with industry, this is really important. It's important to us in terms of our ability to change the world and make the world a better place.

And we know that the impact of our publications is very much higher if we work in a collaborative way. So, there's a thing called the field-weighted citation impact, which is a measure of the impact of your research outputs. And if you have a research output that's only organized by a single author or by a single institution, it has lower impact.

The more collaborative it is, the more likely it is to be cited by others. And that's particularly true for industrial collaboration. So, we know that the field-weighted citation impact goes up if our work is done jointly with industry, because obviously it's then of relevance not just to people in other universities, but it's of relevance to people in industries and way beyond universities. So, I think it's a synergy. I think it's a mutually beneficial relationship. And I think it's an important one.

When it comes to governments, I mean, universities in general in the UK are publicly funded institutions. And so, we're spending public money. So, we have to respect the fact that public money is limited and that there are many calls on the public purse, but we want to be responsible citizens. We want to spend that money wisely. We want to help governments to improve the lives of citizens. And I think universities can make massive contributions to that.

So, I often talk about what I call the other 50%. So roughly 50% of society will understand what a university does. They may have attended university themselves, or their children may, and they have some kind of understanding of what universities do.

The other 50% of society probably don't. And some of them may think universities are not for them or not relevant to them. And we have to be as relevant to that 50% as we are to the 50% that engages directly with us.

Absolutely. Now, I think universities are in a position to make a very major contribution, particularly in this area on AI and its development, but for the betterment one hopes of society as a whole.

But looking at the UK and its international competitiveness, how competitive do you think the UK is in the development of AI compared to other countries? And do you think we're on course to have the necessary infrastructure in place to make the most, if you like, of the enormous opportunity presented by AI?

I strongly hope so. And I think we are in a strong position, but I think we have to be conscious that now we need to sort of run just to keep up, I think, just to stay competitive. I mean, I think the whole world, both universities around the world, but also governments and industries, and indeed the public are sort of alive to AI now.

It's not, there's no kind of privilege to a university like Edinburgh that's been doing it for 60 years. We have to be relevant now and we have to develop into the future. So, I think it's, at the moment, I think the UK is in a good place and I think British universities are in a good place, including Edinburgh, but we can't be complacent about that. We need to actively develop and work with others to understand what the priorities are, otherwise, we will get left behind.

Yeah, absolutely. Now, for our viewers, if you'd like to hear more from Sir Peter and from our other distinguished speakers we have at the upcoming conference, the conference is the National AI Policy, Infrastructure and Skills Summit. It's being held in London on the 12th of November. Further information is available on our website, www.cityandfinancial.com.

Peter, thank you so much for joining us today. Very much looking forward to seeing you in London in November.

Thank you very much for the opportunity, Maurice. And, likewise, I'm looking forward to the conference and from hearing some people who are real experts in the field. Thank you.