The Interview: Bill Gates on AI risk, regulation and the "vacuum of US leadership"

0:00Business is booming. Is your fulfillment ready for peak season? Switch to ShipStation now and handle more volume with confidence when it counts. ShipStation connects Shopify, Amazon, TikTok Shop, and over 200 more integrations with one platform that does it all. You can manage orders, sync inventory, track returns, all automatically. Everything your operation needs, all in one place. Right now, qualified shippers get a real dedicated person to get you set up and switched over for free. So when peak season hits, you're ready to go. Switch today at ShipStation.com. Code audio. This episode of the Times Tech Podcast is sponsored by KPMG.

0:36Katie, we've been tracking the AI surge for a while, but some of the latest data is pretty striking. It is. KPMG's latest quarterly AI pulse suggests that 7 in 10 UK business leaders still see AI as a top investment priority, even if the economy slows down in the next year. But turning that investment into tangible value is another question. And so a lot of organizations are still figuring out how to move from endless AI pilots to something that works across the business. KPMG's focus is on that gap, helping organizations scale from the pilot phase to AI-powered transformation they can trust. To find out more about how KPMG can help your organization make the difference, you can visit kpmg.co.uk forward slash AI.

1:23This episode of the Times Tech Podcast is sponsored by ServiceNow. Danny, one thing we keep hearing from business leaders right now is AI sounds great, but how do you actually make it work inside a company? Exactly, because most organizations aren't neat, shiny systems. They're layers of software, legacy tech, and teams all doing things slightly differently. ServiceNow sits across all that, acting as a control tower for making work move seamlessly through the organization. Connecting people, systems, data, and increasingly AI agents so that things don't happen in silos. Learn how ServiceNow puts AI to work for people at servicenow.com.

2:03This is ultra different than any technology that's come before. And it will change how invention gets done. It'll change life in a huge way. And that's even before the humanoid robots come in and bring the physical capabilities on top of the cognition capabilities. Hello and welcome to the interview from Times Tech. I'm Katie Prescott. And I am Danny Fortson. And this week we have got a pretty big guest, Bill Gates. Yeah, I mean, you can say that again. He's one of the world's leading technologists, one of the world's richest and most powerful people, worth$114 billion, if you must know. And he's been sounding the alarm on AI's potential dangers and the lack of controls to stop them.

2:54He thinks AI is, quote unquote, ultra different from anything that's come before and that governments, universities, and the rest of society just aren't keeping up. And we also, in this conversation, get into the risks to jobs, biosecurity, whether AI is a bubble, and why he doesn't think that the tech companies themselves should be expected to sort all of this out. But he is still hugely optimistic about what it could do for healthcare, for education, and some of the poorest people in the world. So let's get into it. And just a quick reminder, you can watch this episode and all of our episodes on YouTube.

3:35So here's our conversation with Bill Gates. Enjoy. Well, look, there's a huge amount to talk about, and we want to get into the report, all of your very stark warnings that you've issued about AI, and many of the things that you talk about that crop up on our podcast, as you can imagine, week in, week out. But I just wanted to start by looking at the year just gone. You've experienced a huge amount of scrutiny because of your association with Jeffrey Epstein. And I wondered how that was affecting the work that you want to do and what your reflections are on that. I had about a dozen meetings. You know, I was trying to raise billions of dollars for global health and I thought he could refer me to those people and we could save millions of lives.

4:23In retrospect, that was not only a dead end, but a mistake in terms of, you know, the reputational risk. And, you know, we live in a very reputationally sensitive world. You know, I was able to make clear exactly what I was doing in terms of meeting with them and, you know, very clear that that's, you know, all that ever happened there. So, you know, I've learned a lot of lessons over time. You know, I wouldn't even say I had antitrust difficulties. I had a divorce. You know, I'm heavily associated with the tech industry. You know, some things where people question me are things actually that didn't happen, like views of, you know, vaccines tracking people or stuff like that.

5:07But, you know, I'm a lot more careful about who I are the foundation associate with now. And I put a lot of thought into that and doesn't change anything about what the foundation's about or the partners we're working with. Well, thank you for addressing that. And I think, you know, to your point around this report, the world's in a really interesting spot right now. And as a report lays out, especially around AI, there's incredible potential, but also huge negatives, right? And some pretty terrifying directions this can go. do you feel like we're having we're in a window right after which it will be too late like we have to change course we have to take action so that's one question it's two parts the other question is is like of all of the things that could go wrong what is the thing that most worries you that rises above everything else as we think about the arrival of this technology well eventually when the technology gets super good, including being embodied in humanoid robots, you know, perhaps more than 20 years from now, we'll have a world of abundance where food and shelter are provided at either no or extremely low cost because, you know, the robots, the materials, the invention of new medicines, you know, has just gone so far.

6:36And there we'll have challenges in terms of how we organize society and purpose, meaning, things of that nature. That's far enough off and, you know, probably is best thought about by young people. And so a lot of what I'm addressing is this very tumultuous period where AI is starting to affect jobs and creating cyber or bio or psychosocial risk or control risk, that the last year the AIs have gotten powerful enough that those risks are here now. And I know as a close follower and very engaged with using these technologies and consulting with the industry itself, that they're going to get far more powerful even before the end of the year because the hardware and software layers are getting a lot better.

7:29And so I do find myself a bit surprised that the segments of society outside the industry, the universities, the think tanks, both in terms of their use and their sophistication about moderating downsides. They're really not in the conversation. And if we don't have that deep thinking there on things, including how it affects kids or the job market, then you'll have just kind of this black and white, okay, let's ban it all, you know, which then all of the good stuff would be at risk there. And so I'm really surprised it's not more top of mind. And that's, you know, the memo I did last week was to try to raise, you know, say there should be this visibility.

8:23The goalkeepers report is more me, you know, as the full-time chairman of the Gates Foundation talking about how profoundly using AI is influencing our ambition and allowing us, you know, to say, okay, we're going to get the farmers in a better way. We're going to get new medical tools a lot faster. You know, that's mostly on trying to accelerate the positive impacts where a country with, you know, far, far too few doctors, this really is kind of miraculous if it gets implemented and funded the right way. It does feel to me that a lot of what you're saying in your memo, but also in the report, is trying to very publicly shake people out of complacency.

9:08It's like you just said, you know, you're worried it's not top of mind for so many people. But you must have so many conversations behind the scenes with leaders, with government and tech leaders. Do you feel that message is getting through there? No, I'm afraid that a lack of usage, a sense that, hey, past technologies weren't so disruptive and this is just another technology. Just this intuitive sense of, okay, humans are really smart. This thing is going to peak out at some level. You know, I heard it messed up this or that thing. And, you know, maybe that's a permanent flaw that will hold it back.

9:47There's a lot of reasons why the message that this time is different, and this is more profound than any technology, that people would resist that. And, you know, that's why I'm putting all of, you know, my experience and credibility behind the message that it is very different and it requires this broad engagement. I'd say the industry, in terms of at least raising the ideas that there can be harms, I'd say they do pretty well at that. And they alone, though, are not going to self-regulate because there's many of them. And some are in the US, some are in China, some are in other locations. You know, so just when people talk about, well, you know, what are the disagreements in the industry, that doesn't lead to the solution for all of society.

10:43Are you kind of hoping to some degree that the bubble does burst? And what I mean, the bubble, I'm actually talking about, you know, the financial bubble that is being inflated around this. When you talk about the hundreds of billions or trillions of dollars that are just rushing into the industry, a lot of it is debt. We've seen this movie before. Is there a hope on some level that if that blows up, maybe that will be almost like a natural break and give us time to adjust? No, I'm not hoping for some financial thing. And, you know, that's among the topics can distract people. You know, is this stock undervalued or overvalued?

11:23And are we building too few data centers or too many data centers? The AI is in no way a bubble in the sense of like the South Sea bubble or the tulip market bubble. AI is a very, very profound technology. Now, often when you get new technologies, even cars, planes, the internet, you have some investments that you can look back on, say pets.com or something and say, wow, that was pretty excessive of not understanding the competitive dynamics. of what was real value added of the industry. But AI, thinking about what stocks to buy or sell, okay, you can do that. But that's a distraction from the recognition that this is ultra different than any technology that's come before.

12:19And it will change how invention gets done. It'll change life in a huge way. And that's even before the humanoid robots come in and bring the physical capabilities on top of the cognition capabilities that are on this huge ramp. That financial channel won't lead you to, oh, we need to have faster adoption in health and in education. You know, I'm trying to accelerate the positive uses and then minimizing the risk of the bad ones. And the cyber risk is there today. And, you know, even if stocks went down, that doesn't eliminate the cyber risk. It's really striking the report. You talk about increasing adoption, as you just did, and the importance of localized AI.

13:14So how AI in different countries and different cultures needs to be relevant to where it's being used. So AI needs to be in Chichewa in Malawi, not necessarily just in English. But how do you persuade the tech companies that are developing it to do that and sort of pursue good over profit? And should we be doing that? The four companies that the Gates Foundation's engaged with, Google, Microsoft, Anthropic, OpenAI, they're not falling short in terms of wanting to help with beneficial use. You know, OpenAI has this gigantic foundation. Anthropic is helping with a lot of our work. Same thing with Google and Microsoft.

14:02And so, you know, they will provide back-end capacity to the degree we go out and figure out how to get the African languages supported. They'll do that without it necessarily being profitable for them. But they don't understand Africa and African farmers. And so the Gates Foundation, partnering with African governments and other donors, will make sure that the data about health from Africa or the agriculture weather data gets put together in an easy-to-use format. We'll even have benchmarks in health and agriculture to make sure that the local ways things are discussed, the models are very well trained.

14:47So the philanthropic intent of these companies, I'd say, is quite strong. But again, relying on them to understand about kids growing up and how much they should depend on AI and what's tasteful there and not tasteful there. These are not social scientists or psychiatrists, and they will naturally, even though they try to avoid it, they'll bring some industry bias. That is why the economists and social scientists, you know, broadly, the whole of academia should be way more engaged on AI so that eventually you get policy ideas and you get the politicians having a more nuanced view than, hey, I'm for AI or I'm against AI.

15:35On that point, this is a technology that, I mean, as we're seeing with OpenAI and Thropic and just like these kind of unheard of growth rates, it tends to concentrate power. And I'm just wondering what your view is of this kind of generation of tech leaders. Do you think they are ready to meet this moment? In other words, do you trust that this is going to go well? Because as you say, right now, it's a little bit, there's very few guardrails. It's kind of like we're in their hands to a degree. They are not experts on biosecurity. They're hiring in people and all that. But that's not their responsibility.

16:15You know, we have a gigantic government budget and, you know, which has had deep expertise on biosecurity and cyber. And, you know, we really need to activate that. The flaw in this picture is not that the companies aren't concerned or that they're not thoughtful about these issues. And I can only speak to the four that the foundation is deeply engaged in. That's not the flaw. The flaw is that the safeguards, which are going to be democratically determined, but not just a knee-jerk thing, rather deep policy thinkers saying, okay, does the token tax make sense? How would it work? Kids shouldn't be banned from using AI, but what are the boundaries that will make it a net positive where it's helping their curiosity and giving them feedback, but not retarding the challenges of social development?

17:23That's way outside what these companies are focused on. They're going very quickly and amazingly in terms of the depths of the models. And so where is society engaging on this where it's not the role of the industry to self-regulate or understand the whole of society impact that comes out of AI? And you've talked about the creation of some sort of international organization to manage AI's use and monitor AI and the importance of getting the US and China involved in that. Have you got a sense of where the leadership for something like that should come from, how it should emerge? Well, since World War II, the U.S.

18:10has often been a benign leader of global initiatives, whether it's the creation of the UN itself with others, thinking about nuclear weapons or ozone layer or aviation safety. And so here, people see this as important as I do, the capacity of the US government to work domestically and in a friendly way, reach out to others and see what they're doing. It just doesn't feel that's the way the administration is going. The depth of understanding of AI and the right concern about some of these bad things is too low in every entity, including the government, other than the industry itself. There's not many industries that have been like, wow, our tool we'll have to be careful with.

19:05And that's very appropriate here because it is so uniquely powerful. But, you know, OpenAI and even DeepMind talked about, well, if we got to a certain level, we'd have to go beyond profit motives in terms of when and how we release these things. Now, a lot of the design of how that might be exercised was kind of assuming there'd be one company out in front. And so if they managed to hold back, you know, then that would solve the thing. That's not the world we're in. we're in a world with many companies in the United States, many companies in China and some elsewhere. And so it'll have to be the right trade-offs will have to come from outside the industry.

19:50Do you think it's credible though, that for the companies, in a way, it feels like the industry is talking out of both sides of its mouth. It's like, we're building this thing that might destroy humanity. Literally, like there's a non-zero chance. Please stop us. In the meantime, we're going to go as fast as possible. That is a hard thing for people to get their heads around. This is the tool that will help cure cancer or, you know, in the case of the Gates Foundation, help accelerate the eradication of malaria. And so put the industry to the test and say, no, models that, you know, can create new molecules have to be monitored to make sure they're not being used for bioterror applications.

20:34And, you know, where is the credible proposal? It's very hard to say that China won't go along unless we're saying what we want them to go along with. You know, if you're just doing voluntary review, there's nothing to cooperate with. And, you know, I'm saying that you can't go straight to the politicians because it's too complicated. You want the politicians to build up the expertise in the government, but a broad set of institutions have to be developed around these challenges. You know, I had dinner last night with Ginny Raimondo, who's, you know, got a group called Raise that's, you know, raised some money and is going to do some programs to talk about job market adjustment things.

21:20You know, you need a hundred things like that, that eventually leads to new policy. But probably led out of the US, you think? Well, that's been the primary pattern. The rest of the world is not used to having a vacuum of US leadership. You know, we saw that during the pandemic. And, you know, now we have a WHO that doesn't happen to have the United States as a member of it. So no, the world is not fully adjusted to cases where the U.S. isn't in there taking cutting edge issues that need to be solved. And, you know, it's a win-win if we stop cyber attacks. And, you know, I think there's a lot of cooperation that in no way hurts the U.S.

22:12to be working with others. If you look at what's happening now and you go back to your early, early days tinkering with basic all those many, many years ago, are you surprised at where we have got to at the arrival of this technology at this time and how quickly it's developing? You know, throughout my life, the idea of, okay, how do humans do learning and, you know, things like writing really good code, which, you know, is my obsession. The idea of, okay, why can't we understand that and capture that digitally? I've been fascinated by it. And even before I was born, Alan Turing talked about Ed and the term AIs, and it goes way back.

22:53It was about 10 years ago when we started making meaningful progress using these sort of trained statistical systems where we don't even explicitly encode the knowledge. And so it was a little over three years ago when Sam and Greg and Ilya and a group or at my house showing me ChatGPT about six months before the release that I was, you know, shocked and amazed. And then I'd say a comparable experience was late last year as particularly the Claude coding capability, but now not just Claude, passed in many respects my own ability to write code. You know, that to me is like, oh my, you know, that's huge because that announces a threshold of cyber risk and and by risk and so then i expected okay we're going to have everybody coming and talking about this because here we are at these thresholds that we we always knew they were coming but you know i have to say myself i thought they would be come along later and so the the conversation about okay will this eventually make health car better and cheaper?

24:09Will it make housing better and cheaper? Will it make prices go down? How quickly, how can you accelerate those good things? You know, it's interacting with government, you know, now just pleasurable, simple, you know, because AIs are so spectacular at making bureaucracy and complex things work like that. And so the milestones that we've crossed and the lack of broad dialogue that led me to write, you know, kind of in the strongest tone I could, that we're not responding the way that we should be. And that whatever time I get with political leaders or that people pay attention to me, that now has to be a major message for me, along with If they, hey, and don't forget about the poorest in the world who, you know, if we did it right, this will be hugely beneficial for them.

25:10And that, you know, the Gold Keepers report talks about that, you know, we're going to spend over a billion on this and it's infusing all of our strategies. You know, I'm very happy with the way the Gates Foundation is adjusting and thinking to the opportunity. But that's not happening broadly. Right. Well, thank you so much for your time today. It's been great to talk to you. Thank you. Good to talk to you.

26:00scale. KPMG's approach is to look beyond AI as just another technology rollout and instead how it could redesign work across the organization and fit into wider processes and decision making. So it's less about individual use cases and more about how organizations can transform and deliver outcomes powered by AI. To find out more about how it can help your organization make the difference you can visit kpmg.co.uk slash AI. This episode of the Times Tech Podcast is sponsored by health and life insurer Vitality, your health's best friend. Vitality works differently, get active, build healthy habits, and you can unlock rewards from some of the UK's top brands while helping to keep your premiums low.

26:47It's award-winning health and life insurance that helps you live healthier. Find out more at vitality.co.uk. Today's episode of the Times Tech Podcast is sponsored by ServiceNow. ServiceNow describes itself as an AI control tower for business reinvention, connecting people, systems, data, and AI, so workflows more smoothly across an organization. The platform integrates with different clouds, models, and data sources, bringing them together in one place to help your team. To learn more about how ServiceNow puts AI to work for people, visit servicenow.com.

27:28Well, thanks for listening to this episode of The Interview with Bill Gates. We hope you enjoyed it as much as we did. I mean, Danny, it was quite extraordinary to get to talk to him, particularly at this moment, wasn't it? I think the thing that really leapt out to me about it was, and we talk all the time about AI and its importance geopolitically, but it was this vacuum of power, a vacuum of leadership from the US, and the comparison to what happened in the aftermath of the Second World War when these big global bodies were set up, driven by the US. And the fact that he thinks we need one of those, we need something in place in order to manage AI globally, a completely new infrastructure, and that traditionally America makes sense, right, particularly because so many of the models are being developed in the US and so much of the power of AI is coming from the US.

28:17So yeah, to my mind, that was one of the most interesting things, that the world is looking to America for leadership on this and isn't finding it there's a void there's like the opposite of leadership and as we talk about week in week out you know the labs that are building this stuff every week are coming out with a new dire warning which we can talk about like the strategy of that and what they're doing and not doing but what is clear is you know the states is effectively doing nothing and this is where this technology is really being developed and pumped out to the rest of the world and his point that actually it's not the responsibility of companies to grasp the nettle on this that when it comes to things like child safety for example it needs to be governance it needs to be academics thinking about it and actually this lack of understanding as well which i think compounds the lack of leadership or the lack of any the lack of a plan frankly to deal with ai was so interesting to hear from him.

29:20The last thing I would say on that is I think I disagree with that idea that the companies aren't responsible. Because if they believe what they're saying, that they're building something that has a non-zero chance of killing us all, you can't just then say, please slow us down. Please stop us. If those are the stakes, really, and you really believe that, I feel like you really do have to take a different level of responsibility than if you're making whatever m &ms it's never good to start an argument at the end of a podcast is it but i i disagree because i do think it's not right for them to be self-policing somebody has to step in can we really expect these guys in their 30s with the big silicon valley money train pushing them pushing and pushing them into race race to sort of you know faster and faster and develop more and more powerful technology to pull the brakes.

30:18We might want them to, but I agree with Bill Gates that this is bigger than a handful of businesses in the West Coast. Yeah, I was saying that and, you know, rather than that or. But it just feels like sometimes you feel like, well, they're like kind of hands up, like, we can't be stopped, we can't stop ourselves, you know, please help. and you're kind of like, come on, take more of a role here. If you believe what you know your own marketing. It's going to be really interesting to see if his interventions have an impact because he is talking about this very loudly. And it does feel like this sort of the pressure from all sides is ramping up on government.

30:59Certainly, you know, the clamor from the tech industry is getting louder and louder. Obviously, something we'll be returning to a lot on the pod. But please do subscribe if you haven't already. so you never miss an episode like this fabulous chat we just had with Bill Gates. And I think that is it for us this week. See you next week. Bye-bye. This episode of the Times Tech Podcast was sponsored by KPMG. One of the themes we keep coming back to is the gap between AI ambition and real-world application. And that's where KPMG is focusing its work, helping organizations move beyond pilots and think about how AI can change the way the business operates and how it fits into wider processes and decision making.

31:45You can find all their insights and how they can help your organization make the difference through successful AI workforce adoption. Visit kpmg.co.uk slash AI. When you're running a growing business, the right systems can unlock your full sales potential and make fulfilling every order effortless. Switch to ShipStation before peak season so you're prepped and ready before the rush hits. ShipStation connects Shopify, Amazon, TikTok Shop, and over 200 more marketplaces with one platform that does it all. You can manage orders, sync inventory, track returns, all automatically. Everything your operation needs, all in one place.

32:21Upgrade your shipping operations to stress-proof your biggest sales days and capture every bit of potential revenue. Go to shipstation.com and use code audio and get 30 days free. Here's a tip for you. There's a podcast out there with fans waiting to be your next customer. They tune in every week. They trust the host and that host wants to talk about brands like yours in their own words to their audience. The problem is you just haven't been introduced yet. We're a cast where that introduction happens as the world's largest podcast marketplace. We let you browse shows, see who's listening and book host-read sponsorships or run your own ads all from one platform.

33:01Transparent pricing, real-time data, complete control. Start advertising on podcasts by visiting acast.com slash advertise.