Philosophy, AI and the Mess of Leading: the future of institutions in a rapidly changing world | Duncan Ivison

Duncan Ivison, guest · [12:24]There's, there's two kind of general reasons why I think it, it does. The first is that ultimately philosophy is in many ways about understanding the limits of your knowledge. It's about understanding what you don't know and being clear about what you have, as it were, a warrant to believe. That turns out to be an incredibly powerful tool for someone who's in a, who's in a leadership organization like a university where there's so much information coming at you, there are so many different stakeholders, so many different things that you're trying to make sense of across teaching, research, partnerships, you know, the organizational structures, finance, all that stuff. And the discipline and philosophy of, of, of being clear about what the limits of your knowledge are is, is incredibly useful. And I think that leads to a related thing that the boss prepares you very well for, which is how to ask good questions. I think the two are connected. So I think that's really, that's been a really useful set of skills for me that, that I hadn't really reflected on until relatively recently. My management gurus are Aristotle and Kant and John Rawls and Michel Foucault, rather than the traditional business school. Not that there's anything wrong with traditional business schools. I think the second reason philosophy's been such a useful set of skills in retrospect is that foundational first principles thinking is really important the more senior, I guess you get, if I can put it that way, because you want to make sure that the community you're a part of, the organization that you're Helping to lead is oriented around the right goals and values and is focused on the right things. And that requires a certain kind of way of, requires sort of robust first principles thinking. What really matters? What's our mission? And what follows from that in terms of then what we do to, you know, achieve our, our goals, to execute our strategy. And I think it turns out that, at least for me, that philosophical training on, on foundational first principles type thinking combined with that sort of humility about the limits of your knowledge is a good toolkit to start with as a, as a leader when you're making decisions about the organiza, you know, about what to do and what your plans are and what your ambitions are. You're anchoring it in the right way, in the right values, and, and, and you're asking the right questions about whether what you're doing is consistent with those values, and you're being robust about that and not sort of leaving it behind, as it were.

Duncan Ivison, guest · [15:31]And I think that then connects to the whole AI question. I mean, in a nutshell, as extraordinary as generative AI is, large language models don't think. Human beings think. And in an age of AI, the premium will be on thinking and imagination and creativity and human connection and perspective and judgment. And that is the wheelhouse of not just philosophy, but the humanities and arguably, you know, the natural and physical sciences and the other disciplines as well. But In an age of sort of transformational AI, it's human thinking and creativity and imagination and judgment that will matter most and has never before. And I do think that means as much as we need brilliant computer scientists and engineers and machine learning folks, we need people asking the right kind of philosophical questions and ethical questions and value questions about the technology as never before, because that's not going to come from the tech industry. That's not going to come from the people building the machine. It's going to come from people who understand the machines, but are asking very different things of them. And so for me, it's an incredibly exciting time for philosophy. In fact, this weekend I was in Hong Kong with a whole bunch of philosophers and computer scientists, machine learning people, and government officials talking about how it might be possible to build cross-cultural frameworks for governing AI globally, as impossible as that sounds. And, you know, it was incredibly exciting. Like, there's some really interesting work is being done in the field. I think for the young philosophers who were there, who were involved in this project, they're doing really interesting new work. And so I think it's an incredibly exciting time for people in the humanities to be leaning in and contributing to this debate about, you know, how we govern AI, how we regulate AI, how we think about the role of AI in our lives, which is, of course, unavoidable, but we still have choices to make about it. And we have choices about how we use it and employ it and what we ask it to do. And it's really important that we remember that. So I think it's a great, so I think for that reason, philosophy and the social sciences, the humanities in general, it, it, it's really a golden era in many ways for, for, for these disciplines, despite the fact it might feel the opposite.

Duncan Ivison, guest · [22:21]So, so we decided that aligned with our strategy, we would basically find a way to make the technology available to everyone in our community for free without restriction. And we struck a partnership with Microsoft, first of its kind, I think, in the world. So, but for now, I'm sure others will catch up, that sees all 65,000 people in our community get access to the, the latest, you know, version of, of, of Copilot. But at the same time, a whole range of training and support, and also a framework for testing, challenging, and debating the technology. So alongside of providing access to students and, and staff, And faculty to, to use the technology, we've set up a whole series of working groups that are looking at the environmental impact of generative AI, the ethics and responsible use of generative AI, the, the way in which we might deploy it internally in, within the organization and how we should do that. I see it as a kind of living natural experiment about how we can use this transformative technology, in the right way. Because I don't think the tech companies are going to do it. I don't think governments are going to do it for us. I think we've got the opportunity to do it for ourselves. I think it's, we, we have a responsibility, I think, to do it. The other thing is we want our students to go out into the world with the appropriate kind of skills and, and capabilities, but also the right kind of ethical framework in a world that is going to be completely transformed by AI. I mean, I feel it would be irresponsible not to prepare our students for that world in this way.

Duncan Ivison, guest · [24:51]Now, having said all that, it's been quite controversial, right? Which is, I kind of welcome. I mean, you know, the first, Town hall, I do regular town halls, and the first town hall I had after we announced the Microsoft partnership, I think we had 1,700 people, you know, rock up, and it ranged from, this is evil, and you're destroying the institution, through to, where's my license? So I, you know, it's like, and I really welcome that because it is disruptive. It is discombobulating. It is challenging. We don't have all the answers. We don't know everything about the, the ultimate impact and consequences of this technology. But I believe strongly that it, it's at universities that we should be asking these questions and giving people the chance to challenge and deploy and use and experiment with this technologies. So yeah, we've really, we've really kind of decided to go for it, lean into it, give people access, but also have the debate. And, you know, we were really clear we wanted to make sure that it shouldn't depend on whether you, you buy it or you have to borrow a license off your friend. We also wanted to create a safe environment in which people can use the technology so that the beauty of the partnership with Microsoft is we have a really secure environment within which people can use the technology in, in all the appropriate ways and in all the ways that matter for an organization like us. So we're on our way, but it's been, you know, I really welcome the, the, the debate and the conversation that's followed in the wake of it.

Duncan Ivison, guest · [27:58]Huge amount of learnings because I think I underestimated the emotion about AI. I mean, I knew it was, obviously I knew it was a big issue for all of us, but I think I underestimated or, or I've learned a lot about just how passionate people are. About the risks of AI and the sense of not being in control of what the technology might do to us as an organization and to them as, you know, individuals, whether students or colleagues or partners. That was a real learning. You know, it was something that we probably should have anticipated, but, and it's been in a way a good process, but I was like, wow, we wanted to move fast because we had an opportunity and things are moving so fast. We, we didn't want to, you know, wait 18 months and have a hundred committees. We decided to go for it. And what I learned was there's a lot of anxiety and, and a lot of emotion wrapped up in, in what the people feel this technology is doing to our culture, to our, to our families, to, to our communities. Some good, but a lot bad. And, you know, so that was a really, that was a good learning, I think, moment for, for me in particular. It made me really want to make sure we were listening and, and, and adapting and adjusting and, and, and, you know, then adapting our, our, our engagement and our comms around this whole program of work in the right, in the appropriate way. Well,

Daniel Atlin, host · [29:34]I imagine that engagement and dialogue is, is a huge part of this journey, actually, and, and creating forums that you can actually, you know, cultivate that as well.

Duncan Ivison, guest · [29:42]Yeah, absolutely. It's really important. You know, there's a, there's a concept in political philosophy called deliberated disagreement, and it sounds very Hokie and very, you know, so liberal and everything, but I really believe in it. There's a big difference between when you and I disagree without actually talking to each other, we just make assumptions about each other's beliefs and say, oh, that guy's a jerk. There's a big difference between that and actually having a conversation, and we might still disagree just as passionately. At the end of that conversation, but something has happened between us. There's been some recognition of, you know, between us. And I think that really matters in a university environment, which is fused with disagreement. So in my experience, turning up and having the conversation, standing up at the front of the room or whatever and saying, okay, let's talk about this, doesn't solve every problem, doesn't mean people leave feeling the world is instead of right. But there's a big difference between doing that and not doing it and just saying, well, we disagree, so we just leave it at that. You can't do it for everything, but for the things that really matter, it's important.

Duncan Ivison, guest · [31:44]I think we need to organize ourselves more around the problems and challenges that the world faces. Rather than expecting the world to organize itself around our disciplines. And that doesn't mean abandoning disciplines, it doesn't mean, because you need the depth and the robustness and the richness of disciplinary training and knowledge to do great science, to do great literature and, and philosophy and history. But I think the universities of, of the great universities of the 21st century are going to look and feel different in that sense. And moreover, they're going to give our students more of that opportunity to see problems and questions in the realm and to have the skills to be able to do that more importantly. So yeah, I guess if I was, could wave a magic wand and say, let's refound the University of Manchester, I think we might have Faculties that look very different and schools that look very different, and we'd be more porous in terms of the kinds of institutions, organizations we worked with, and our students would have a, a more porous kind of experience of their own learning. And I think, I think increasingly our social license to operate will increasingly rely on us being able to organize ourselves in those ways. I think, I think the community really respects the deep discovery research we do. They really respect the, the quality of, of the teaching that my colleagues do, but at the same time, they, they increasingly want the university to help figure stuff out that will make their lives better. And I know for some of my colleagues, that's like anathema, but I think It, it, it's something we need to really embrace and make sense of and start to organize ourselves around.

Duncan Ivison, guest · [37:03]I mean, the first thing is, although, you know, the buck stops with you when you're in these roles, and you always need to remember that, you have to have, you have to focus on your team. You, you have to build a great team. And I certainly made lots of rookie mistakes in various roles in the past where I didn't really understand that. I didn't really understand that profoundly enough. So the first thing is, You need to build a team. You need to be in a situation where you're making decisions collectively, and you're drawing on the talents and expertise of the people around the table, around the organization that can help you get to the right answer and get to the right decision. So I think for me, the first thing is, you know, there are many aspects of the job that are kind of lonely, if I could put it that way, where it really is you that has to make the call, and often it's between two bad options rather than a good option and a bad option. But it's also important to, to create an environment in which you're, you're making decisions collectively, you're, you're working as part of a team. And in my experience, the best universities have very strong cultures of distributed leadership in that sense. And I think really university presidents who I've really admired and learned from have been able to get that balance right between Giving people a sense of empowerment and allowing those who are closest to the action to make the decisions in the best interest of the organization, but then also knowing when to step in and say, look, no, this is the direction we need to go in, which means we need to make this call and this is the way it's going to go. So I think for me, the first thing is you're, you're building a good team, you're, you're investing in that distributed leadership model, but then you're crystal clear about your role and, and the unavoidability of having to sometimes make a decision that will be deeply unpopular, but you feel is the right one for the organization. You know, there are moments when nobody wants to make the call because it's too hard, too painful, too difficult, and that's your, you've got to take on that responsibility. You can't, you can't sort of chuck it over to a committee, you know. But if you get it right, you're, you're in a dialogue with people about The organization, you've got the information and, and inputs that enable you to, to, to make the best possible decision in the circumstances.