S10E4: The AI Risk Debate - Hype or Existential Threat?

SPEAKER_1 (0:00)

Welcome to Age of AI, Politics, Law and Society, hosted by Medlir Mema, Chris Lamont, and Young Diogenes in association with the Global Governance Institute in Brussels.

SPEAKER_2 (0:20)

Imagine walking up one morning to find that your smart assistant, once just a helpful tool for setting reminders, has begun making decisions on your behalf. Filtering your news, prioritizing your emails, even scheduling appointments without your approval. At first, it's convenient. Then, it's unsettling. Who's really in control? Or picture a world where AI surpasses human intelligence, not with malice, but with indifference. Optimizing the economy, the environment, even human behavior in ways we never anticipated. Will we thrive in a world designed by machines, or would we find ourselves obsolete? These are just glimpses of the debate around AI risk. Some argue that AI is humanity's greatest invention, a force for innovation, efficiency, and progress. And others warn that unchecked, it could pose an existential threat. Not because it hates us, but because it might not need us. So how real are the risks? Is AI an ally or a silent adversary? To make sense of these questions and more, I'm delighted to have with me today, Torben Swoboda and Risto Uuk. Torben, Risto, welcome to the Age of AI.

Torben Swoboda (1:31)

Thank you so much for having us.

Risto Uuk (1:33)

Thank you. Thank you. Good to be here.

SPEAKER_2 (1:35)

Well, of course, the pretext we're having you here is a fantastic short paper that you wrote, a very interesting take on AI risk and arguments for and against it and primarily kind of evaluating. And we'll work through that paper together because that is really the base of our conversation. But before we do that, how about a short introduction on both of you? Maybe Torben, we start with you and then Risto, your turn.

Torben Swoboda (1:58)

Sure. Yeah. So my name is Torben Swoboda. I'm a doctoral student at KU Leuven in Belgium. I'm also part of the Leuven Business School, where I'm part of the Technology and Innovation Cluster. And in my research, I'm primarily focused on AI safety and ethical AI, and all with an interest in AI governance, and with a background in philosophy and economics. So I'm very interdisciplinary researcher and that has been very fruitful in the domain of AI governance.

SPEAKER_2 (2:33)

Thank you. Risto.

Risto Uuk (2:36)

I work as the head of EU Policy and Research at the Future of Life Institute here in Brussels, and I've been working at FLI for the last four years, almost. I lead our policy and research efforts, especially in relation to the EU AI governance efforts. For example, the EU AI Act, which was passed quite recently, and now is entering into the implementation phase. So I lead our efforts here. And I'm also a PhD researcher at KU Leuven, where I study systemic risks from general purpose AI. So what kinds of risks are we talking about? How could we assess them? How could we mitigate them? Those kinds of things. Since we were talking about interdisciplinary backgrounds, I also have quite different things that I've studied. I did philosophy and economics in my undergraduate, then I did philosophy and public policy.

At the London School of Economics in my master's. And now I have a general social science discipline, which is partly in philosophy, partly in economics, sociology, AI, whatever the field that might be particularly relevant at particular moments.

Torben Swoboda (3:46)

That's what I do.

SPEAKER_2 (3:48)

Now before we get into the heart of the matter, we have just come out on the other side of a very busy few days, few weeks, when it comes to discussions on the future of AI.

Between the AI Action Summit in Paris, the NATO Defense Ministerial in Brussels, and the Munich Security Conference, there's a lot to talk about and to think about when it comes to AI. There's a lot to think about other things as well. We will not talk about those geopolitical conversations that are happening in Europe and on the other side of the Atlantic. But just focusing on the AI itself, what are your key takeaways from the AI Action Summit in particular and other events? And how do you see the shift away from safety in favor of action? There used to be AI Safety Summit, now it's AI Action Summit. What's your take on that? Risto, maybe we can start with you and Torben, follow up.

Risto Uuk (4:37)

Yeah, I just came back from Paris. I was there for a whole week. I attended lots of site events, lots of meetings. Future of Life Institute's president, Max Stegmark, was at the actual Action Summit. So I could hear some takes from him, but also in media, from collaborators, other stakeholders. And I think the Action Summit could have easily been called the French Investment Summit or something like that, because it was essentially France trying to raise investment nationally and hype their own technological development capabilities, and even to the point of trying to claim that they are part of an AI race now. And really, it was very major shifts from what happened in the UK and South Korea, which was a lot of focus on governing the advanced AI systems for the benefit of humanity from the perspective of risk, safety, and this one really sidetracked all of those conversations about safety. They were not part of it. I would maybe point out that one key issue at this particular summit was that right before the summit, one of the leading computer scientists of our time, Atresu Abenjio, who I think currently is the most well-cited, most highly cited computer scientist in the whole world, he published a report on the state of AI safety.