WEBVTT - The $2 Trillion Global Arms Race 0:00:02.720 --> 0:00:07.240 Bloomberg Audio Studios, Podcasts, radio news. 0:00:17.360 --> 0:00:20.560 I'm Stephanie Flanders, head of Government and Economics at Bloomberg, 0:00:20.760 --> 0:00:23.239 and this is Trumpnomics, the podcast that looks at the 0:00:23.280 --> 0:00:27.360 economic world of Donald Trump, how he's already shaking up 0:00:27.360 --> 0:00:29.720 the global economy and what on earth is going to 0:00:29.720 --> 0:00:35.360 happen next. This week, we have NATO leaders, donald Trump 0:00:35.360 --> 0:00:38.319 among them, gathering in Turkey for what could be one 0:00:38.360 --> 0:00:43.479 of the alliance's more consequential summits. I'm recording on Tuesday morning, 0:00:43.800 --> 0:00:46.839 US time, and President Donald Trump has already arrived in Anchoror. 0:00:47.240 --> 0:00:50.280 He's expected to meet with the Turkish president there and 0:00:50.479 --> 0:00:54.880 other leaders. And the focus of these meetings recently, particularly 0:00:55.320 --> 0:00:58.520 under President Trump, has tended to be money. The constant 0:00:58.560 --> 0:01:01.840 and reasonable refriras from the US being that Europe needs 0:01:01.840 --> 0:01:04.759 to spend more on confronting the risks at their borders 0:01:05.040 --> 0:01:08.360 and shouldn't continue to rely only on the security blanket 0:01:08.640 --> 0:01:12.280 of the US. But it's not just NATO. Defense spending 0:01:12.400 --> 0:01:15.479 is ramping up all over the world in a new 0:01:15.840 --> 0:01:20.319 global arms race that could look and feel very different 0:01:20.560 --> 0:01:22.480 from the last one. For the first time that's the 0:01:22.520 --> 0:01:26.160 Cold War. Many countries are aiming to acquire nuclear weapons, 0:01:26.280 --> 0:01:29.520 or at least look like they might want to, and 0:01:29.560 --> 0:01:33.399 the guardrails previously constraining that proliferation have more or less 0:01:33.680 --> 0:01:38.440 fallen away. We're also seeing very dramatically how drones and 0:01:38.600 --> 0:01:42.280 AI have transformed the battlefield in Ukraine and Iran, and 0:01:42.319 --> 0:01:45.400 that has countries trying to beat their rivals to the 0:01:45.480 --> 0:01:49.440 next generation of hypersonic weapons and other things. So all 0:01:49.480 --> 0:01:52.000 of this, I think could raise big questions not just 0:01:52.080 --> 0:01:54.640 about the conduct of future wars, but also given the 0:01:54.640 --> 0:01:59.000 amounts involved shape parts of our economy. So what are 0:01:59.040 --> 0:02:02.120 the key features of this new arms race and what's 0:02:02.160 --> 0:02:04.720 it going to mean for the global economy of governments 0:02:05.120 --> 0:02:08.520 one after the other make military tech one of their 0:02:08.560 --> 0:02:13.400 biggest investment priorities. Becker Wasser is the Defense lead for 0:02:13.400 --> 0:02:16.720 Bloomberg Economics. Previously, she led studies and wargames for the 0:02:16.800 --> 0:02:19.640 US Department of Defense at the Center for a New 0:02:19.680 --> 0:02:22.960 American Security and that the RAND Corporation. Becker has also 0:02:23.000 --> 0:02:26.360 been a staff member on a congressionally appointed Committee on 0:02:26.480 --> 0:02:29.799 the National Defense Strategy. Becker, welcome back to Trumpennomics. 0:02:29.919 --> 0:02:30.880 Yeah, thanks for having me. 0:02:31.320 --> 0:02:35.400 And Jerry Doyle is Global Defense editor on the news 0:02:35.440 --> 0:02:38.560 side for Bloomberg News. Jerry, thanks for coming, Thanks for 0:02:38.560 --> 0:02:47.280 having me. Jerry, just tell us what are people buying 0:02:47.360 --> 0:02:50.639 or what are government's buying and why is so much 0:02:50.680 --> 0:02:51.760 of it happening right now. 0:02:52.440 --> 0:02:56.320 Well, so much of this investment in defense technology kind 0:02:56.320 --> 0:02:59.680 of started with the Russian invasion of Ukraine. That's sort 0:02:59.680 --> 0:03:03.239 of urgency to the whole situation, not just in Europe 0:03:03.240 --> 0:03:07.480 but globally, as countries started to quickly realize that they 0:03:07.560 --> 0:03:10.120 might have underinvested since the end of the Cold War. 0:03:10.639 --> 0:03:13.240 So that led a lot of country to start buying 0:03:13.320 --> 0:03:15.280 a lot of the traditional things that you expect to 0:03:15.320 --> 0:03:18.840 see on a battlefield, the tanks, the planes, missiles, artillery 0:03:18.880 --> 0:03:21.639 with a huge priority. And now we've sort of move 0:03:21.760 --> 0:03:24.679 past that and started to look at what is going 0:03:24.720 --> 0:03:27.919 to be the technology that's going to dominate the future 0:03:27.960 --> 0:03:32.160 battlefield for decades. During the Cold War, you saw that 0:03:32.280 --> 0:03:35.520 manifest itself. That arms race manifests itself in terms of 0:03:35.600 --> 0:03:39.920 nuclear weapons, not just the total stockpile sizes, but you know, 0:03:40.040 --> 0:03:43.120 strange novel waves of delivering them, different effects that you 0:03:43.160 --> 0:03:47.080 can create with with nuclear weapons. And this current arms 0:03:47.160 --> 0:03:49.640 race that we're seeing right now is not about that 0:03:49.720 --> 0:03:52.839 it's not about, you know, being able to deliver megatons 0:03:52.880 --> 0:03:56.280 with the death onto a city. It's sort of more subtle. 0:03:56.360 --> 0:03:59.520 It's trying to find ways to give yourself an advantage 0:03:59.520 --> 0:04:03.120 in the future, doing things like investment drones for different 0:04:03.160 --> 0:04:07.000 types of things, for defense, for offense, for surveillance, the 0:04:07.000 --> 0:04:11.160 militarization of space, trying to figure out ways to protect 0:04:11.200 --> 0:04:14.440 your satellites in orbit, trying to find ways to give 0:04:14.480 --> 0:04:17.440 your military an advantage with the things that you place 0:04:17.520 --> 0:04:20.839 in orbit. And then hypersonic weapons, which are sort of 0:04:21.160 --> 0:04:25.279 a more traditional means of destruction but with a novel 0:04:25.360 --> 0:04:29.600 twist of being able to maneuverate extremely high feeds, avoid 0:04:29.800 --> 0:04:33.279 air defenses and so on. And we've seen that, and 0:04:33.320 --> 0:04:37.159 we've seen the drones sort of in practice happening on 0:04:37.160 --> 0:04:39.520 battlefield right now, and it's about trying to figure out 0:04:39.520 --> 0:04:41.760 a way to find that, trying to figure out ways 0:04:41.800 --> 0:04:46.200 to improve and apply these things on a large scale. 0:04:46.360 --> 0:04:49.040 I think Jerry rays is something that's really interesting, and 0:04:49.080 --> 0:04:53.119 it's using all of these technologies to gain an advantage. 0:04:53.279 --> 0:04:55.200 But what does that look like? And I think here 0:04:55.320 --> 0:04:58.640 there's a few things. There's the decision advantage. Who can 0:04:58.680 --> 0:05:01.880 make decisions faster and therefore gain the initiative, and that's 0:05:01.880 --> 0:05:05.400 where things like AI enabled technologies come in. But you 0:05:05.560 --> 0:05:09.200 also have this idea of mass, who can put more 0:05:09.279 --> 0:05:12.880 on the battlefield at any given time to overwhelm defenses, 0:05:13.240 --> 0:05:16.440 to overwhelm your adversary, and that's where you see things 0:05:16.520 --> 0:05:20.279 like low cost drones and lower cost missiles come in. 0:05:20.800 --> 0:05:23.200 And then I think the last area of advantage is 0:05:23.240 --> 0:05:27.920 this idea of cost. Who can do that cheapest? And 0:05:27.960 --> 0:05:31.560 that's where some of these more traditional weapons systems, these 0:05:31.640 --> 0:05:36.159 higher end missiles, hypersonics space, that's all very expensive. And 0:05:36.200 --> 0:05:38.560 that's why you see a lot of governments and a 0:05:38.560 --> 0:05:41.839 lot of militaries going all in on AI and all 0:05:41.880 --> 0:05:44.800 in on drones because they think that it can get 0:05:44.839 --> 0:05:48.559 them the best effect for the cheapest bang on the buck. 0:05:49.120 --> 0:05:51.800 But just on their own, drones aren't going to be 0:05:51.880 --> 0:05:55.000 the solution to all problems, and I think sometimes we 0:05:55.080 --> 0:05:58.600 do see a lot of militaries, particularly those that either 0:05:58.640 --> 0:06:03.120 have manpower constraint like Ukraine, or those who have fiscal 0:06:03.160 --> 0:06:06.960 constraints like arguably some countries in Europe, going all in 0:06:07.000 --> 0:06:10.080 on drones to try and make up for those sharkfalls. 0:06:09.880 --> 0:06:14.000 It's absolutely true that you know, drones are a useful tool, right, 0:06:14.040 --> 0:06:17.520 but they don't really fully replace a lot of the 0:06:17.960 --> 0:06:21.719 high end weapons. You talk about destroying a tank with 0:06:21.800 --> 0:06:25.159 these tiny first person view drones that you see in Ukraine, 0:06:25.520 --> 0:06:28.400 what we see in the social media is the drone 0:06:28.480 --> 0:06:30.320 hitting the tank and the tank blowing up. We don't 0:06:30.360 --> 0:06:34.000 see like the previous twelve to fifteen first person drones 0:06:34.000 --> 0:06:36.520 that hit the tank and did nothing or disabled it, 0:06:36.880 --> 0:06:38.680 or the fact that the tank had already run over 0:06:38.680 --> 0:06:42.240 a mine and was stationary. Right, So the effects that 0:06:42.279 --> 0:06:45.720 we get from these cheap weapons aren't necessarily replacing what's there. 0:06:45.760 --> 0:06:49.800 They become an additional tool on top of what already exists. 0:06:50.040 --> 0:06:52.160 We'll get into some of the sort of numbers involved 0:06:52.760 --> 0:06:55.200 in a minute. And also I think the way it 0:06:55.200 --> 0:06:59.880 could shape economies. But as someone who's not as fully 0:07:00.400 --> 0:07:03.640 absorbing this stuff the way you both are every day, 0:07:04.160 --> 0:07:06.240 there seem to be there's sort of two things that 0:07:06.279 --> 0:07:08.280 we get told, and it feels like they can't both 0:07:08.480 --> 0:07:11.640 be true. On the one hand, at least in the 0:07:11.640 --> 0:07:14.000 early stages of the Ukraine War and people learning about 0:07:14.080 --> 0:07:16.480 drones and drones becoming a much bigger element of all 0:07:16.520 --> 0:07:19.000 of these battlefields. We got the sense that it was 0:07:19.040 --> 0:07:21.680 going to be cheap, and as Becky you say, you know, 0:07:21.720 --> 0:07:24.080 you think of it as going to be good, high 0:07:24.160 --> 0:07:26.800 bang for your buck. But there's also this sense of 0:07:27.120 --> 0:07:30.640 the sheer amount of technology involved in getting the most 0:07:30.640 --> 0:07:32.800 out of them and also developing all these other kinds 0:07:32.800 --> 0:07:35.640 of arms suggests that, you know, none of this is 0:07:35.640 --> 0:07:38.280 getting cheaper, and if anything, it's still giving a massive 0:07:38.320 --> 0:07:40.800 advantage to the countries that have the deepest pockets. There 0:07:40.840 --> 0:07:42.280 seem to be different strands. It's kind of on the 0:07:42.320 --> 0:07:44.440 one hand, you talk about it in some of your 0:07:44.560 --> 0:07:48.440 stuff about being a leveling technology AI and some of 0:07:48.440 --> 0:07:51.480 the other modern technology in innovation, you can have a 0:07:51.520 --> 0:07:55.280 country that's much smaller Ukraine's a classolic example, kind of 0:07:55.280 --> 0:07:57.920 punch above its weight. But on the other hand, it 0:07:57.960 --> 0:08:00.800 does seem that scale is still incredibly important for being 0:08:00.840 --> 0:08:03.320 able to move ahead with these cussing head technologies. 0:08:03.840 --> 0:08:06.880 And I think it's fair to say that a country 0:08:06.920 --> 0:08:09.920 with deeper pockets like a China or the United States 0:08:10.200 --> 0:08:12.640 would be able to do more with that. They would 0:08:12.640 --> 0:08:15.000 be able to put more resources into production, be able 0:08:15.000 --> 0:08:18.040 to put more resources into research and development, and so on. 0:08:18.640 --> 0:08:22.080 But the effect that they're able to get in terms 0:08:22.160 --> 0:08:25.600 of magnifying the power of their own military is going 0:08:25.680 --> 0:08:28.640 to be diminished because they have all these other capabilities 0:08:28.680 --> 0:08:31.080 that kind of do the same thing or do it better. 0:08:31.560 --> 0:08:34.720 And so what drones and some of these other technologies 0:08:34.760 --> 0:08:37.240 you mentioned AI is another good example. What they do 0:08:37.520 --> 0:08:42.360 is enable somebody who can't afford the high end, maybe 0:08:42.400 --> 0:08:45.679 who can't afford a bunch of six generation fighters, who 0:08:45.720 --> 0:08:49.719 doesn't have massive constellations of satellites overhead, who doesn't have 0:08:49.840 --> 0:08:52.840 cruise missiles that can fly thousands of miles, they're going 0:08:52.920 --> 0:08:56.240 to be able to get something like eighty percent of 0:08:56.320 --> 0:08:59.720 the combat effectiveness, being able to do eighty percent of 0:08:59.760 --> 0:09:03.240 the thing for a lot less money. And if you 0:09:03.280 --> 0:09:06.280 are a resource constrained country or military, that's a good 0:09:06.320 --> 0:09:06.800 trade off. 0:09:07.440 --> 0:09:09.319 We should sort of get a sense of some of 0:09:09.400 --> 0:09:14.240 the numbers involved when I talk about how those countries 0:09:14.280 --> 0:09:16.520 are ramping up spending around the world, I mean, what 0:09:16.600 --> 0:09:18.520 kind of numbers are we talking about for the increase 0:09:18.559 --> 0:09:19.439 in defense spending. 0:09:20.000 --> 0:09:21.839 Well, if you want to look at just the US 0:09:21.880 --> 0:09:26.000 system alone, the US defense budget requests this year is 0:09:26.080 --> 0:09:29.720 one point five trillion dollars. That's just for this year, 0:09:29.880 --> 0:09:33.200 and that amounts to roughly three point four percent of 0:09:33.320 --> 0:09:38.199 GDP spent on defense for the US. That's actually sizeably 0:09:38.360 --> 0:09:41.480 lower than what it had been spending in the Cold War, 0:09:41.760 --> 0:09:45.719 where it had a much higher percentage as a shriff GDP. 0:09:45.559 --> 0:09:48.120 A lot higher than it was a few years ago exactly. 0:09:48.400 --> 0:09:50.400 And I mean, I think again, if we want to 0:09:50.440 --> 0:09:52.800 pull it back to the NATO summit, we're defense spending 0:09:52.960 --> 0:09:56.280 is very much going to be the topic of conversation. 0:09:56.840 --> 0:09:59.200 You do see the ramp up in Europe to move 0:09:59.360 --> 0:10:02.600 toward that of three point five percent of GDP to 0:10:02.640 --> 0:10:06.800 be spent on core defense spending by twenty thirty five 0:10:06.840 --> 0:10:08.839 and twenty thirty some of the work that we've done 0:10:08.840 --> 0:10:11.800 at Bloomberg Economics show that, you know, the top countries 0:10:11.840 --> 0:10:14.960 in Europe are likely to undershoot that target and that 0:10:15.080 --> 0:10:17.360 it's going to be a very slow ramp up to 0:10:17.600 --> 0:10:20.800 get to three point five percent of GDP, and that's 0:10:20.840 --> 0:10:24.520 also going to come at a really high amount of 0:10:24.880 --> 0:10:25.640 taking on debt. 0:10:26.480 --> 0:10:28.760 Jerry, I think the figure that we had for the 0:10:28.800 --> 0:10:31.400 sort of global piece of this, at least in some aspects, 0:10:31.440 --> 0:10:37.200 was two trillion dollars across three continents, which is not 0:10:37.640 --> 0:10:39.840 far off the sort of some of the numbers we 0:10:39.960 --> 0:10:44.680 used to in relative terms for the Cold War nuclear 0:10:44.840 --> 0:10:47.559 build ups. One thing we know from that research that 0:10:47.880 --> 0:10:51.079 Becka mentioned is that the any kind of economic dividend, 0:10:51.160 --> 0:10:52.800 you know, government's like to talk about, well, and we're 0:10:52.840 --> 0:10:55.400 spending more and it's going to produce jobs and manufacturing, 0:10:55.440 --> 0:10:57.480 and it all sounds great if you can get past 0:10:57.520 --> 0:11:02.120 the military spending part, but when you actually look at 0:11:02.160 --> 0:11:06.480 the how much economic impact there is, it's probably less 0:11:06.520 --> 0:11:10.600 than other kinds of infrastructure spending. And in Europe particularly, 0:11:10.600 --> 0:11:12.560 it's going to depend a lot on whether they can 0:11:12.600 --> 0:11:16.520 grow the domestic defense sector because otherwise they're going to 0:11:16.559 --> 0:11:19.160 just be bringing in a lot of imports from the US. 0:11:19.200 --> 0:11:20.360 So how are they looking on that? 0:11:20.880 --> 0:11:26.320 Individually? European countries are not integrate position like. Many of 0:11:26.360 --> 0:11:30.400 them have a defense champion of various size, but usually 0:11:30.400 --> 0:11:34.040 there's a focus on one thing. They make avion X, 0:11:34.160 --> 0:11:36.960 or they make ships, or they make air defense system, 0:11:37.040 --> 0:11:41.040 they make tanks, arty shells, engines, something like that. Rarely 0:11:41.160 --> 0:11:43.400 in a case sort of like Germany or France, where 0:11:43.520 --> 0:11:47.280 you have even bigger companies or an assortment of companies 0:11:47.320 --> 0:11:49.079 that can do a lot of these things at a 0:11:49.120 --> 0:11:51.560 high level and sort of enable them to have a 0:11:51.679 --> 0:11:54.760 very well developed defense industry. But in order for Europe 0:11:54.800 --> 0:11:57.199 to really compete in this global arms rates that we're 0:11:57.240 --> 0:12:00.320 talking about, there has to be collective actions. There have 0:12:00.360 --> 0:12:03.960 to be coordination about who's building what. There has to 0:12:04.000 --> 0:12:07.360 be coordination on spending, There has to be coordination on 0:12:07.400 --> 0:12:10.480 what the priorities are, which right now we see as 0:12:10.600 --> 0:12:15.440 being again satellites, space based military infrastructure. We see air 0:12:15.480 --> 0:12:18.560 defense as being another big one, and then there's sort 0:12:18.559 --> 0:12:23.199 of more boring but high impact stuff like transport aircraft, 0:12:23.320 --> 0:12:26.160 like strategic airlift we call it, being able to move 0:12:26.280 --> 0:12:29.400 tanks and people all around the continent pretty easily. That's 0:12:29.440 --> 0:12:31.679 something that Europe doesn't have a lot of, and it 0:12:31.760 --> 0:12:35.440 kind of relies on the US for on NATO tanker 0:12:35.520 --> 0:12:39.240 aircraft that refuel other planes in flight and extending their range. 0:12:39.280 --> 0:12:41.400 That's another thing they lean heavily on the US for 0:12:42.120 --> 0:12:45.320 and airborne warning and control, which is something that the 0:12:45.440 --> 0:12:48.600 US again has a lot of and historically has been 0:12:48.640 --> 0:12:52.320 sort of generous about sharing with NATO partnerships, but that 0:12:52.720 --> 0:12:55.800 Europe is very interested in doing for themselves right now, 0:12:56.080 --> 0:13:00.200 and we saw that announced with the GLOBALI aircraft. They're 0:13:00.200 --> 0:13:02.720 going to buy ten of those for NATO to be 0:13:02.880 --> 0:13:05.600 airborne wanting control plane that are going to replace the 0:13:06.040 --> 0:13:09.960 older American made planes. So they're moving in that direction 0:13:10.040 --> 0:13:12.000 and it's important for a lot of reasons. 0:13:11.960 --> 0:13:16.080 Right now Europe in terms of defense spending, in terms 0:13:16.120 --> 0:13:20.839 of translating that into genuine capabilities and who's producing those capabilities. 0:13:21.240 --> 0:13:24.960 It's incredibly fragmented, and you have a fragmented defense industrial 0:13:25.000 --> 0:13:28.600 base across Europe, and eventually there's going to need to 0:13:28.640 --> 0:13:32.480 be this requirement to pick winners and losers, but that 0:13:32.640 --> 0:13:37.000 is politically unpalatable right now because of the economics, because 0:13:37.160 --> 0:13:39.840 every single country wants to see that increase in defense 0:13:39.880 --> 0:13:42.680 spend go back to their own economy, go back to 0:13:42.720 --> 0:13:45.880 their local economies, and be able to have a conversation 0:13:46.360 --> 0:13:49.960 that's no longer sort of this like guns versus butter debate. 0:13:50.520 --> 0:13:52.640 But to be fair, that's always been an element of 0:13:52.679 --> 0:13:54.800 the US debate as well when it comes to states. 0:13:54.960 --> 0:13:57.600 Everybody wants their own piece, and that has also meant 0:13:57.600 --> 0:14:00.199 that often defense spending in the US was incredibly an. 0:14:00.000 --> 0:14:04.640 Efficient, incredibly and very much concentrated in certain pockets where 0:14:04.679 --> 0:14:08.000 you had certain members of Congress who have continually fought 0:14:08.240 --> 0:14:12.040 for specific systems to not be cut when perhaps maybe 0:14:12.080 --> 0:14:14.760 it was time for them to retire because of the 0:14:14.760 --> 0:14:18.400 benefits that they brought to their district or their state. 0:14:30.400 --> 0:14:33.120 Jerry had sort of started off talking to us about 0:14:33.560 --> 0:14:36.360 the sort of tech piece of this and how that 0:14:36.440 --> 0:14:39.120 was driving some of these investments. And if we sort 0:14:39.120 --> 0:14:41.560 of combine that with the discussion we've just had, I mean, 0:14:41.600 --> 0:14:45.840 it does suggest that if governments are going to be 0:14:46.360 --> 0:14:50.760 investing more in research and development in the military space 0:14:51.240 --> 0:14:55.120 and needing to be much more connected to the companies 0:14:55.400 --> 0:14:59.960 that are developing these new technologies, that feels like even 0:15:00.440 --> 0:15:03.280 the future of AI itself, well, certainly whole areas of 0:15:03.320 --> 0:15:05.920 technology could be shaped by the fact that we're in 0:15:05.960 --> 0:15:09.640 a time of globally increasing defense spending. 0:15:10.520 --> 0:15:12.920 Yeah, I think so. And you know, if you look 0:15:12.960 --> 0:15:16.720 at a country like China, where there's this intense interweaving 0:15:17.000 --> 0:15:19.840 of the public and the private sector and you know, 0:15:19.880 --> 0:15:22.200 the universities and the government. 0:15:22.400 --> 0:15:24.760 You talk about the civil military fusion. 0:15:24.560 --> 0:15:28.360 Exactly exactly where you know, in some sense that gives 0:15:28.480 --> 0:15:33.400 China an advantage because any any breakthrough that happens, whether 0:15:33.440 --> 0:15:36.520 it's at a private company or a research university, or 0:15:36.600 --> 0:15:39.640 you know in some you know, random lap in the hinterlands, 0:15:39.680 --> 0:15:42.120 it will end up in the hands of the Chinese government, 0:15:42.560 --> 0:15:45.080 or they will at least have the opportunity to evaluate 0:15:45.080 --> 0:15:47.120 it and decide whether that's going to be you know, 0:15:47.240 --> 0:15:53.600 militarily applicable or not. That also translates into traditional defense 0:15:53.600 --> 0:15:56.440 buildouts as well. You know, the Chinese government decides that 0:15:56.480 --> 0:15:59.560 they need more aircraft, there's not really a fight over 0:15:59.640 --> 0:16:02.000 you know, whose congressional district they get built in or 0:16:02.000 --> 0:16:04.800 anything like that. You know, sort of the machine starts up, 0:16:05.760 --> 0:16:07.560 and then you compare and contrast that to the way 0:16:07.600 --> 0:16:10.400 Europe is set up, whereas Beca said, there's always going 0:16:10.440 --> 0:16:12.040 to be friction about, you know, who's going to be 0:16:12.040 --> 0:16:14.080 the winner, who's going to be the loser. We saw that, 0:16:14.400 --> 0:16:17.280 you know, f cast program, which was supposed to you know, 0:16:17.360 --> 0:16:22.400 have the two largest aircraft companies in Europe, like the 0:16:22.520 --> 0:16:25.680 so and Airbus are going to combine forces and make 0:16:25.720 --> 0:16:28.680 the sixth generation plane. We saw that fall apart because 0:16:28.680 --> 0:16:31.520 the two companies couldn't get along, and that's not something 0:16:31.920 --> 0:16:36.040 really that happens in China, so to that extent they 0:16:36.160 --> 0:16:38.200 might have an advantage and the way they moved this 0:16:38.280 --> 0:16:39.560 sort of innovation. 0:16:39.240 --> 0:16:42.560 Forward, Becca, I'm sort of HARKing back on this point 0:16:42.600 --> 0:16:47.280 about the closer connection between the government and the tech sector. 0:16:47.760 --> 0:16:49.680 I mean, I guess we saw that most recently with 0:16:49.720 --> 0:16:52.520 that between Anthropic and the Pentagon. A lot of people 0:16:52.560 --> 0:16:55.800 would be concerned that the biggest tech firms who are 0:16:55.840 --> 0:16:58.520 right at the forefront of this and the biggest amounts 0:16:58.520 --> 0:17:02.120 of money are going into the military applications of AI 0:17:02.640 --> 0:17:04.639 rather than other things. Is that I mean, should we 0:17:04.720 --> 0:17:06.560 be worried that that's just going to shape the future 0:17:06.560 --> 0:17:07.040 of AI? 0:17:07.480 --> 0:17:10.160 I mean we should be worried about, you know, how 0:17:10.240 --> 0:17:13.200 AI is going to be integrated into the military. More broadly, 0:17:13.520 --> 0:17:16.760 there are some very benign applications. One of the ones 0:17:16.760 --> 0:17:19.359 that I've heard the most from the US military is 0:17:19.400 --> 0:17:23.920 trying to use AI to speed up essentially, it's modeling 0:17:24.040 --> 0:17:28.000 for how it gets logistics from one place to another. Right, 0:17:28.119 --> 0:17:31.200 that's to me, that's a benign way, Like, Okay, let's 0:17:31.240 --> 0:17:34.600 speed up processes. Let's speed up things that take tons 0:17:34.600 --> 0:17:37.480 of hours of people just you know, hand jamming things 0:17:37.480 --> 0:17:42.200 into a spreadsheet like it's nineteen ninety nine instead, you know, 0:17:42.320 --> 0:17:44.879 trying to think about the ways in which you can 0:17:44.960 --> 0:17:47.600 have good use cases, but you could also equally have 0:17:47.880 --> 0:17:51.440 some negative use cases. And traditionally the Pentagon has put 0:17:51.480 --> 0:17:55.080 some guardrails. Some of those guardrails have actually been relaxed 0:17:55.560 --> 0:17:59.880 by the Trump administration. So now we've seen in places 0:17:59.920 --> 0:18:03.320 on the battlefield, specifically in the Iran War, where there's 0:18:03.320 --> 0:18:07.560 been heavy use of AI tools, particularly Pallinteers madeven smart 0:18:07.600 --> 0:18:10.320 system to try and speed up some of the targeting 0:18:10.359 --> 0:18:14.440 processes to get that decision making advantage that so many 0:18:14.480 --> 0:18:18.520 militaries want. You know, that puts the US and frankly 0:18:18.600 --> 0:18:22.159 other militaries on a potentially tricky slope. That's not to 0:18:22.200 --> 0:18:24.800 say that there are not guardrails against using AI for 0:18:24.880 --> 0:18:28.000 some of the most dangerous things. I think there's a 0:18:28.160 --> 0:18:31.879 lot of pushback, and rightfully so, on the integration of 0:18:32.040 --> 0:18:37.920 AI into nuclear command and control. So essentially, whether AI 0:18:38.000 --> 0:18:41.840 can help make decisions about the use or non use 0:18:41.960 --> 0:18:45.520 of nuclear weapons, to me, that should be very much 0:18:45.680 --> 0:18:47.800 a human decision making process. 0:18:48.240 --> 0:18:52.160 Jerry I started mentioning about the nuclear piece of countries 0:18:52.240 --> 0:18:54.560 might have concluded from the way the world has gone 0:18:54.640 --> 0:18:59.280 recently that actually they have an interest in acquiring nuclear weapons. Supposedly, 0:18:59.280 --> 0:19:01.840 we only have nine nuclear powers currently in the world. 0:19:01.920 --> 0:19:06.800 But are we actually seeing countries developing an increase in 0:19:06.800 --> 0:19:09.080 the number of countries wanting to develop and is there 0:19:09.080 --> 0:19:12.119 a potential and rather frightening overlap with our previous conversation 0:19:12.200 --> 0:19:13.120 about AI. 0:19:13.320 --> 0:19:16.600 In the defense world, there's always the possibility of frightening overlap. 0:19:17.400 --> 0:19:20.760 It's frightening overlap all the way down. But looking at 0:19:20.840 --> 0:19:24.960 nuclear weapons specifically, as far as we know, nobody who 0:19:25.080 --> 0:19:28.760 doesn't currently have nuclear weapons has actively spun up a 0:19:28.800 --> 0:19:32.440 program to create them. What we have seen is expression 0:19:32.480 --> 0:19:36.720 of interest from European countries about getting more deeply involved 0:19:36.760 --> 0:19:41.600 with existing nuclear programs. Specifically, Francis macrong has spoken, you know, 0:19:41.680 --> 0:19:44.480 sort of in very plain terms about France wanting to 0:19:44.520 --> 0:19:48.000 expand it's arsenal, wanting to offer a nuclear umbrella to 0:19:48.040 --> 0:19:51.240 the rest of Europe to sort of replace what he sees, 0:19:51.280 --> 0:19:54.320 as you know, the US pulling back from its commitments there. 0:19:54.840 --> 0:19:58.119 We also see in Asia sort of open conversations in 0:19:58.200 --> 0:20:02.120 Japan and South Korea about acquiring nuclear weapons that work 0:20:02.160 --> 0:20:05.240 hasn't started yet as far as we've known, and there 0:20:05.240 --> 0:20:08.600 hasn't been any official government decision to move that forward. 0:20:08.640 --> 0:20:10.960 But the fact that those conversations are happening in the 0:20:11.000 --> 0:20:14.320 open says it is real right that desire, that fear 0:20:14.440 --> 0:20:17.560 is there. Nobody wants to get left behind. And when 0:20:17.560 --> 0:20:21.440 you see what's happened in places like Ukraine and Iran, 0:20:21.960 --> 0:20:25.959 the easiest equation to apply to that is, well, you know, 0:20:26.040 --> 0:20:29.000 if those countries had had nuclear weapons, like North Korea, 0:20:29.359 --> 0:20:34.080 this wouldn't have happened. That's exactly right. And as Becka said, further, 0:20:34.200 --> 0:20:39.760 applying developing technologies to nuclear arsenal things like artificial intelligence, 0:20:40.080 --> 0:20:42.920 that is a concern and it's something to be worried about. 0:20:42.960 --> 0:20:46.800 I think at this point everybody who has nuclear weapons 0:20:47.000 --> 0:20:51.160 is extremely aware of the problems that they could come 0:20:51.240 --> 0:20:55.840 from linking artificial intelligence to nuclear stockpiles. When I was 0:20:55.920 --> 0:20:59.040 in Singapore for the Shngrol dialogues, there's a higher ranking 0:20:59.160 --> 0:21:03.880 generals in both the People's Liberation Army China and Pakisthan 0:21:04.000 --> 0:21:07.720 military sort of spoke openly about this that their fear 0:21:07.840 --> 0:21:10.000 of AI was not that it would sort of you know, 0:21:10.440 --> 0:21:13.680 take over in sort of an actual artificial intelligence way, 0:21:14.040 --> 0:21:18.520 but that tools using AI would speed up decision making 0:21:18.560 --> 0:21:22.760 processes to the point where people would sort of was 0:21:22.800 --> 0:21:25.280 the cool Essentially, you know, they would think, you know, 0:21:25.320 --> 0:21:28.160 instead of having twenty minutes to respond to a nuclear 0:21:28.200 --> 0:21:31.159 attack or ten minutes or something, everything would move so 0:21:31.359 --> 0:21:35.960 fast with you know, recommendation for responses and things like that, 0:21:35.960 --> 0:21:38.240 that somebody would make a rast decision and things would 0:21:38.240 --> 0:21:40.920 get worse instead of better. So those are the kind 0:21:40.920 --> 0:21:44.919 of guardrails I think that broadly countries are looking to 0:21:45.080 --> 0:21:46.159 apply to that sort of thing. 0:21:46.280 --> 0:21:48.200 And of course we did have examples in the Cold 0:21:48.200 --> 0:21:50.520 War of that, you know, whether the machine or is 0:21:50.560 --> 0:21:54.840 telling an individual pilot or whatever this is this is 0:21:54.880 --> 0:21:57.639 a nuclear attack, and at least there were minutes there 0:21:57.720 --> 0:22:00.520 as opposed to seconds for them to think is this realistic? 0:22:00.800 --> 0:22:04.120 And it reglied on humans ultimately to make the decision 0:22:04.160 --> 0:22:05.399 about what to do next. 0:22:06.240 --> 0:22:09.760 In the past, defense departments have been very good at 0:22:09.800 --> 0:22:13.120 talking a good game about fantastic new technologies that were 0:22:13.160 --> 0:22:15.720 coming down the track. We had Star Wars, now we've 0:22:15.720 --> 0:22:19.400 had I mean, Becker has done some very elaborate costing 0:22:19.920 --> 0:22:22.040 that where the numbers get very large for the so 0:22:22.160 --> 0:22:25.399 called Golden Dome that Donald Trump has talked about for 0:22:25.440 --> 0:22:28.239 the whole of the US, and an awful loss of 0:22:28.280 --> 0:22:31.359 that money has been wasted. They talked a good game 0:22:31.400 --> 0:22:34.119 but didn't really pan out. Is an awful lot of 0:22:34.119 --> 0:22:35.479 this money just going to be wasted? 0:22:36.119 --> 0:22:40.880 I mean, yes, It's hard to say how much, though, 0:22:40.920 --> 0:22:44.920