Deeply Seeking - China’s AI Challenge

[00:00:00] The Why Curve, with Phil Dobbie and Roger Hearing. It's cheap and it's from China. And it's as good or better than what Western tech has built. The DeepSeek artificial intelligence app has upended the prospects for AI. And how you make it. And it crashed the shares of the Silicon Valley giants. So has China, with its multitude of tech scientists, found a breakthrough that the West couldn't? Or has it just stolen and repurposed what ChatGPT and the others were already doing?

[00:00:30] And how will all this change the future of AI, the dominant technology of the age? The why. I have to say even before all of this story broke, there was the question, wasn't there, about whether AI had been overblown. Yes. A bit like the dot-com bomb. Were we piling money into something, which I mean there's no doubt there's something there. And a lot of people are finding it very useful. But are the productivity gains that we're going to get from it as big as, you know, the money would suggest? Exactly.

[00:00:58] Which is part of the reason I think that all the tech firms had such a drop when DeepSeek turned up. Because I think they suddenly thought, oh my God, we didn't need to put all this money in because actually this is cheap and it works better. Yeah. But is it still doing the same job? I mean, are the benefits, you know, there's still a question over the benefits out of it all. Well, there is. But I think the feeling is that it's, you know, the case for the necessity of it and the effectiveness is pretty strong. I mean, all the big things that Apple's building have got AI built into them. And unless they, I mean, that's one thing.

[00:01:25] If they can find a cheap mechanism to use it, it'll help them an awful lot. Do you use it day to day? I don't. Do you? Well, it's hard when you're a journalist, isn't it? Because you actually are supposed to be researching and generating content yourself. It's sort of like an admission of defeat if you tend to. Oh, I bet there are plenty who are using it. I'm sure there are. Absolutely. And you can see some benefits for it. But like maybe you've got to in these days, this same age, you've got to repurpose content in lots of different ways. So actually having artificial intelligence just rewording stuff, maybe there's nothing wrong with that. Well, there isn't.

[00:01:55] But a lot of it, don't forget, is also gives you the wrong information. Yeah. Like there isn't enough wrong information out there already. Well, I, a while ago, I did a, you know, a check on who I was. I said, who's Phil Dobby into an AI? They said, no idea. Yeah. Well, it came back and basically gave me a profile of your life. Oh, well, thank God for that. Yeah. So that helped you enormously. Well, I started to question my achievements. Yeah. Might get you a job though. Yeah, mine. Who knows? Yeah, yeah. It would happen. No, no. I mean, I think this is the problem. It's a fallacious system, basically.

[00:02:25] Well, let's look more at it today with our guest. Yeah. Who's joining us? Dr. Daniela Dalvia, Senior Lecturer in Banking and Finance Law at Queen Mary University of London, who joins us now. So, Daniele, I mean, is this the news about DeepSeek as big as it seemed to be initially, that China has come up with something that is better than Silicon Valley, was working on, costing less? I mean, have they basically just undercut those, those Californian giants? Well, it's a good question.

[00:02:54] I think that actually DeepSeek is a new financial innovation in the artificial intelligence. But it's also true that possibly this is coming even in a critical moment where Trump, as we know, and the new administration in the US was treated in China about imposing new tariffs. So essentially, it seems to me that the movement probably is not my chance to widespread the news about this.

[00:03:22] And also, as you've seen, I mean, in the press, this is something that we still have to understand how the artificial intelligence is working within DeepSeek. Because I was even reading recently, for example, Italy stopped and blocked the availability of DeepSeek in its own Google application. So there are lots of suspicions about what it is, because same, I suppose, with other Chinese products,

[00:03:49] I mean, Huawei and things like that, where a feeling that if it gets into our communication system, that it may be being used for other purposes in Beijing. Yes. Essentially, I mean, there is a private concern, because essentially there are the form of the state is completely different from our democracies in the West. And so, for example, in Italy, the regulator of the privacy stopped anything until they don't have further information,

[00:04:16] how in reality the information that is shared on the platform is used for which purposes. So on these points of view, as you know, European Union is very thorough on regulations of privacy, and the same also in the US. So we have essentially, we are aligned, especially in G7 countries. So I think that this is something that is important to think about also, not only if there is a new financial innovation,

[00:04:42] how good it is, but also which value it provides and also how the information that we share are used. But isn't it, I thought the idea was that it was, it was like an open architecture that you could take the DeepSeek model, you could sit it on a server in Milan if you wanted to, and use it as a learning algorithm, you know,

[00:05:08] feed it whatever you wanted to create your own artificial intelligence community, you know, without any connections back to China. Yes, essentially, let's say that in financial innovations, everything is about coding and testing. So essentially, if there is someone that is introducing a new idea, another one can copy that idea and can make it better. So this is something that even in financial history, we have seen many times.

[00:05:35] But the main key question is, this is working or not? So it's something that can create maybe a bubble like we saw in the dot com bubble in the 90s. But it's something that probably can be different. I think that every time there is, there is a financial innovation, there is also a rush from investors. So everyone would like to invest and to move those money into the next big thing.

[00:06:01] And on the other hand, I mean, states, they try to regain power to claim that they are maybe the only one producing the best artificial intelligence product. So that is something that goes between geopolitics and power, the relations and the movement. Well, do you think it is a bubble then, Daniele? Do you think you think this is just something that is the latest fad or is it a real step change?

[00:06:29] Because people are saying it's so cheap, it's easy to produce. And that that's something which China has advanced on and the West hasn't. Yeah, I mean, there were many rumours, as we've seen also in Bloomberg and other financial newspapers, think that China was copying already models that were developed by Chagipity, for example, even the last product that was the 4.0. But it seems that DeepSeq in some way can even be better than the 4.0 product.

[00:06:58] If it's a bubble. And that is that that's a question for investors, isn't it? Who've invested in Silicon Valley with high costs, with very expensive Nvidia chips that are very hungry for power. You know, they consume a great deal. Here's something which is less power hungry. Has it been developed for a fraction of the cost? You'd have to question your investment in big tech, wouldn't you? Yeah, essentially, let's say, I mean, Silicon Valley is famous for this.

[00:07:27] So essentially, every time that there is a new innovation is always coming in our society from Silicon Valley. But it's also true that this is something that, I mean, everything, every time has to be tested, because investors basically sometimes they back on companies. So even though maybe the company today is not having a real product or is not even producing, if we think about electric vehicles, for example,

[00:07:56] sometimes I am specialized in special purpose acquisition companies. And one, for example, sector of special purpose acquisition companies that was very famous was electric vehicles. But sometimes those companies that they were bought, they were not even producing electric vehicles. So this is really showing how there is an interest of investors sometimes on betting to the future on something that is not even there. Right. Brochureware. Betting on brochureware, basically.

[00:08:26] It's the power of the PowerPoint presentation. But this is something that is working. I mean, people have got DeepSeq. They put it on their computers or their smartphones. They've used it. They've produced whatever it is. More downloads in the Apple store in the UK. And it works. So it's not a phantom, is it? Absolutely. I mean, this is how in reality, financial markets, they work and how financial innovations also are produced.

[00:08:52] How in reality, investors, they switch different risk or they diversify their portfolio of investments. And that is how even in financial histories, it's created actually financial innovations and sustained our growth in our economy.

[00:09:12] So when the DeepSeq news broke and, you know, it emerged that it was, you know, such a popular download and then, you know, stories about, well, it's actually able to go, you know, to a deeper level than the existing search engines for a fraction of the price. The share market took a hit. Big tech, you know, saw prices falling quite considerably, but only for a day or two. And then it all bounced back. So what happened there?

[00:09:39] Is it just the power of the Magnificent Seven just cannot be beaten, irrespective of what happens? Yes. Let's say the power of the Magnificent Seven, I mean, was under threat for sure after that news. And so that news, I think that is important to show that there might be a systemic effect.

[00:09:58] And this systemic effect sometimes can be detrimental, but it's also true that essentially New York Stock Exchange, NASDAQ or exchanges in New York are the most liquid in the world. And still today are the place where investors or companies, they want to get listed. So this is very difficult to change. So essentially the supremacy of the US in securities is uncontested.

[00:10:27] So this will make no difference, the whole thing, the tech jump up and down. It's yesterday's story. We move on. Or has this actually made a fundamental difference? I think it will make no difference in the sense that we will not take off the supremacy of the US or will not take off possibly also the importance of NVIDIA. That is the other company in the US that was hit by the news. But you're saying for systematic reasons or systemic reasons.

[00:10:57] I'm just wondering why. If there's a product which is better that doesn't need those NVIDIA chips. And it's cheaper. And it's cheaper. Why would the markets? I mean, okay, people say, well, you know, the American market is supreme. We've got a lot of money in US dollars that are sitting there in the share market. We want to concentrate where the growth is. So let's do it in the United States. We can't even invest in this deep seek company elsewhere. So let's just put our money where we can.

[00:11:24] I mean, you'd still have question marks, wouldn't you, about how safe that money is investing in big tech, given it's got someone, you know, nipping at its heels in quite a big way. Yes, essentially, it's more to diversify the portfolio. So in reality, I mean, not all the money are taken away from NVIDIA. But generally, an investor can even edge its own bets in this way if it's going to diversify risk.

[00:11:52] But if you did that, sorry to interrupt, but if you did that, if you were diversifying your risk and moving away from these companies, then surely we'd see a more sustained reduction in the share prices. And we're not seeing that. Yes, at the start there was. But then after, I mean, there were many comments also on the security, on the privacy security, data protection security systems of this other application. So the investors, they started to be worried.

[00:12:21] Essentially, every rumor in the market can change the chart in some way. But still, I think it's more the symbolic effect of this DeepSeq to show that essentially everyone else can have this product at the low cost. And also if you think about Brazil, for example, my article in the conversation was recently also translated in Portuguese,

[00:12:43] because Brazil actually is thinking to develop a new model on the base of DeepSeq, for example. So to give a possibility not only to the rich, but to give a possibility to everyone to share a platform where artificial intelligence can be used. So that's really interesting. So you're saying Brazil is going to develop something similar to DeepSeq or will they just use DeepSeq? Yes, use DeepSeq as their basis. Yes, that's the basis. And why can't others?

[00:13:13] Why can't Europe do that? Because it seems that the big companies that have invested heavily in artificial intelligence are also looking at vertical integration. So they are buying up data centers to run their artificial intelligence. They're doing everything they can to almost close the loop so that they strengthen their stronghold. And yet here's something coming along saying, well, you don't need to do that.

[00:13:36] Stick DeepSeq in a data center, wherever it might be in Europe or whatever, and you can get what you're getting from California cheaper and have more control over it without being tied to a, you know, a very inclusive sort of closed garden model, which they like so much on the West Coast of America. Yes, I think that, I mean, all these questions that relate also to the regulation that we have, for example, in Europe is quite complex.

[00:14:02] So essentially all those kinds of concentration of companies that they buy each other or they consolidate their power, they can do a vertical merger, horizontal merger, but at the end of the day, this has to be approved by the European Commission, for example. So we live in an environment probably in the European Union, but also similar in the UK because it's always a G7 country, the US where antitrust is very important.

[00:14:31] So in reality, all these processes are stopped also at some point. While in other countries where China, in reality, the state is in control or other countries that they go towards a model of deregulation, because let's say also Trump would like to have deregulation, but we all know that he cannot have it at all. I mean, all 100%.

[00:14:55] And this is why possibly, I mean, other countries with less regulation and with more ground to develop those financial innovations and this kind of mergers also that they can happen as you said. There may be, it's a fertile ground where you can start to develop. The problem is that probably you cannot reach a global audience at some point. Yeah.

[00:15:21] There will always be problems because such we have seen very recently in Italy, the regulator has stopped the use of DeepSeq. Yeah. And Daniel, what about the whole relationship between China and the US? Because one of the things people said about DeepSeq's arrival was America has been trying to stop the access that China would have to microchips, microprocessors, and that this would be a good way of holding them back. Does DeepSeq tell us that actually that's just not possible, that actually they can get around all this? Yeah, yeah.

[00:15:50] And in fact, did it help them? The fact that they couldn't get access to NVIDIA chips, so they needed to develop something that used less power, you know, perhaps gave them a better result at the end of the day. Does it change the balance, if you like, between China and the US? Yeah, essentially. I mean, I think that this is like a sort of competition between the two. So at some point it would have happened in any case. I don't think that it's only based on that.

[00:16:15] I think that essentially the competition was always open, probably with a new administration in the US became even more compelling for Chinese to be even more competitive. And I think this issue of competition is around global financial markets. So today also, if you think about the UK, it's no more part of the European Union.

[00:16:38] And in fact, all the new legislations or even, I mean, the new government is promoting a competition, regulation by competition. So to become more competitive, to boost the production of the country and to in reality lure the attention of investors in your own system and country. And the same essential is a little bit everywhere in the world today. Of course.

[00:17:05] It can be China, can be any also developing country like Brazil that now after the Zika is trying to achieve the same results. And they're not concerned about the China connection because presumably, because they're going to be running their own instance of it. So they're not. And doesn't it, once you start doing that, the whole argument about security from China sort of disappears out the window of it, doesn't it? You're just using the code that has been developed, which is open source code.

[00:17:34] Anyone can get access to it. That's what I find really strange out of all of this. It's open source code. So there's billions. It's not making me plenty money who originated it then either, is it? Well, I mean, they make money. So people who develop open source, they make money on how you adapt the open source or what you stick on top of it. You know, it's because they understand the code. But it's, but I mean, I still don't get the security argument when we are talking about open source code.

[00:18:01] Yeah, essentially, I mean, security argument is if we even use ChatGPT. Every time we are trading, we are trading open AI. So in reality, we are providing information to the software. Sometimes we are also sharing private information. We don't know also how users, for example, are using those tools and also for which purposes.

[00:18:24] And this is something that today, at the moment, there are not even general principles that are agreed at international level. And this fragmentation of regulations that they will come always more in the future. That is what can even create regulatory arbitrage. So as soon as there will be a state that maybe is more severe in the regulatory environment,

[00:18:52] the producer or in reality, I mean, the startups that will move somewhere else. So I don't think that the environment we are living is very positive even to sustain them in time such financial innovation. I think that we need to aim for an international cooperation among countries. That's not very likely, though, is it really? Not currently, it would seem.

[00:19:20] I mean, the world is moving in the opposite direction, if anything, isn't it? The idea that Europe says, hey, let's regulate some American companies a bit more. I can't see that's going to go very well with Donald Trump, is it? I mean, or even, you know, even on shore in America, the idea of more anti-trust legislation is not going to go down very well with the president. Yes, exactly. I mean, I said that this moment is the opposite. I hope that essentially in the future can be on the other side of the argument or more into cooperation,

[00:19:49] because essentially that worked very well for banking, for example. If we think in the 1970s, I mean, once we didn't have still the Battle Committee on Banking Supervision, and then with the advent of the Battle Committee on Banking Supervision, we could have harmonization within banking law regulations and how even to protect customers or to mitigate financial crises.

[00:20:18] So this is a process that essentially will possibly follow the same path in artificial intelligence. But it's more complicated, isn't it? I mean, banking, I know, banking is very complicated as well. I think even people within banks don't necessarily quite understand how the banking sector works, but at least it's sort of relatively static. I mean, here we are talking about an evolving technology which is touching all aspects of society, so it's very difficult to really understand how you legislate.

[00:20:48] Absolutely. This is a good point because this story about NVIDIA and DeepSeek actually is touching all our lives. So it's something that is not only related maybe to a large language model, but it's really related to self-driving, to healthcare, to security. It can be essentially, I mean, everything. And essentially, Daniel, it's beyond, in a way, what could be controlled.

[00:21:16] This is one of the arguments about AI, and there's something called AGI, Advanced General Intelligence, where it learns to such an extent that you're not actually really aware of how it works, but also not really in control of it either because it is self-generating. And that's going to be immensely hard to legislate about or control in any way. But also, the companies, does it get back to the same companies that are providing that capability? In which case, do we end up, you know, there's four companies in the world,

[00:21:44] and, well, you know, Magnificent Seven becomes the Magnificent Four, and they're doing almost everything. I mean, again, it gets back to that antitrust legislation, which will need to be powerful to stop this concentration of power the more we are reliant on a core technology, if you like, which is touching so many aspects of our lives. Absolutely. And I think also that the new future will be synthetic intelligence, so not even artificial intelligence.

[00:22:13] So it's really one that we will plug into reality, the visual intelligence. And so this is something that essentially, I mean, they are also inventing some microchips that you can essentially then connect to your brain. And that, I think, will be a new revolution even. Yes, that's rather scary, that one. Is China going to be a big player in this, though? Because if they, it seems like any advance they make,

[00:22:43] the idea of something that you put in your brain, and it's made in China, for most people who... Is alarm bells going off for you? Well, for a lot of people, there would be. But, I mean, perhaps unwarranted. I mean, the idea that, you know, something that is made by Jeff Bezos that you stick in your brain, you'd be concerned about. You know, you'd think, hang on, my Amazon account, I'm sure I didn't order that. But I just thought about it. I mean, it's, you know, it's just how does that, how do we keep control out of all of that? But also, are we just always going to say, we don't trust China,

[00:23:11] therefore they may have great ideas, they may have great technology, but they can't bring it to market, because outside China, there's this huge mistrust, along with America also, you know, obviously trying to build up their mistrust, because they want to sell their own stuff. Absolutely. I think, I mean, here is an issue of democracy. So essentially, even if we think about central bank digital currencies today, that is a big topic, for example, the European Union wants to adopt a digital euro, UK a digital pound.

[00:23:41] US for sure now is not into the digital dollar. But let's say this topic will have a lot of difficulty to be conceived in China, because many people, they will always see that that is an instrument of control. How I spend my money, this time will be seen by a central bank. And so in China... I'm sorry, but I think the same about Europe and the UK, by the way.

[00:24:08] I just feel like the tax officers are going to be onto it straight away, aren't they? But there's a difference here, Daniel, about China, because China has the great firewall of China, whatever. They are not part of the international internet, the World Wide Web. They're just not. They're separate from it. So whatever happens with AI out there, in a way, they are sealed off from it. They don't have to accept any of this that we think about what might happen. Yes.

[00:24:34] I mean, as I said, the reality, then the power to become global will be more difficult always, because we don't accept them in any case. So there will be always an ardor from the European Union, an ardor from the US and all the G7 countries, that they can cooperate or coordinate the report together in case, not to adopt the technology, for example. And so that technology will be only seen in China.

[00:25:03] But China is giving, in reality, an hope to other developing countries, like Brazil, like South America, Africa, where in reality, I mean, even there are many people, they don't even have access to a bank account. And so to see that in reality, something can be produced at a low cost and then be available to many. And so maybe to be able to copy and paste that code and maybe to make your own code on that application,

[00:25:32] this is something that possibly can be appealing, so can give democracy within the accessibility of a product. That's really interesting. DeepSeek being the democratization of AI in that sense. And doesn't that create a situation then where you've got all of these developing countries that are moving faster down the AI road because they've got a cheaper implementation, whereas in the West we're going, no, don't trust China, let's stick with the expensive stuff

[00:25:59] they're making out of Silicon Valley. I mean, we could be putting ourselves at a disadvantage, couldn't we? And this is a very good argument. Essentially, we don't know how the future can look for sure, but there is a trend. I think I see a trend towards that path, where probably if that is the trend, Trump is right to go into the deregulation,

[00:26:20] the deregulation of AI in the US and try to really give any possibility for startups to develop products without... Without competition, yeah. But could that competition include DeepSeek? DeepSeek? No, I don't think so because it's... But an implementation of DeepSeek that isn't tied back, you know, sort of like an open source instance of DeepSeek, not a problem with that?

[00:26:46] I mean, I think in all these things, one point is the ethical concern also. So there are not still very clear ethical guidelines. And so that possibly will be a first step where if there are clear ethical guidelines, we will also understand how to use those instruments and how they can be used. But what about the economic implications of this? Because if, as you say, there's a kind of democratization of AI,

[00:27:14] that it gets out there into developing countries, they can do their own thing with it. Won't this actually generate quite a lot of global economic movement? And in places where perhaps it hasn't been before, and actually that will be a terrifically good thing. Until the economic balance perhaps slightly more in favor of these developing countries. The point of developing countries is in any case, I mean, the infrastructures. So, I mean, we need also to see in reality once they develop that technology,

[00:27:41] what they can develop with which level of even how they can export that technology also in other countries. Because essentially, but it's also true that possibly it's good for their own purposes, to give access to their own cities and to give a product for their own citizens. Yes, I mean, I think that would be the main advantage for them. For sure not to become leader in the market.

[00:28:09] That is something that I always see that has always been between, in case China or the US, big powers. But outside those areas, I mean, I'm thinking of something, for example, like the development of smart mobile technology hugely transformed economies in Africa because people in the middle of nowhere could find out market prices and things. So something of that order. It also became a bank for them. Yes, exactly. So maybe that, Daniele, should be the way things go forward.

[00:28:36] Yes, that, for example, if you think about Kenya and PESA, that is an alternative means of payment that has consented to all the people there to have access to payment systems. And that is the problem was because there was not a bank. So really, I mean, there was not access to a banking system for those citizens. And so we invented alternatives of payment systems.

[00:29:03] I think, of course, financial evolutions or financial innovations, I think are useful for this reason to give, in reality, democracy to people. The point is that then, I mean, those financial innovations, they must be regulated. There must be some limit within them. Otherwise, they might be used also for negative, I mean, poor. Well, that's inevitable, isn't it? I mean, you know, the regulation is not going to be that effective

[00:29:30] in those sort of places necessarily because they're simply not the force to implement. But irrespective of where in the world you're talking about, I mean, all the investment so far has been in companies, you know, investors have been putting money into the magnificent seven or, well, five or six when it looks at AI. And the money's just been there in building the AI. I'm yet to see money going into industries that could benefit from AI, as though people aren't really quite sure what those industries are going to be.

[00:29:59] They've got this wonderful, shiny new thing, but we don't know what to do with it. Exactly. So at what point do we actually turn around and say, because I think that this is an important question, because then it becomes the, well, you know, if this is the opportunity, then we need, you know, whatever technology is going to best serve it at the lowest price and focus less on those companies and focus on what the benefits it's going to provide. But I'm not hearing too much. Are you about, you know, which companies are going to do really well out of AI in the future?

[00:30:26] And certainly, you know, we're not seeing investment dollars going to those companies. Yeah. Also, I mean, it's not only in Vedia and DeepSeq. Like, even there is Alibaba, for example, in China that is creating or trying to create a new product in that sphere, in that sector. And also investors, they are looking at Alibaba, for example. Then there is Meta. I mean, there are so many players that they are coming to the space.

[00:30:54] But is it just the online providers and their interaction with their customers? Or is it going to be good for medicine and make, you know, in which case, why are we not seeing that? Pharmaceutical companies. Yeah. Or exploration for mineral resources or whatever, or retail, the retail sector. For example, let's say healthcare is something that, I mean, AI in general is already 10 years that we have it. So it's not something completely new that happened today. Something that they started to use it also in healthcare.

[00:31:23] And the problem is that in healthcare, even there, there is the privacy of the patients, for example. So especially in the European Union, but also the same is in the UK. All these systems that are centralized, that are very useful because, I don't know, they can then generate all your medical history. The problem is privacy because if someone can breach that system, can take all data, personal data.

[00:31:50] Which brings us back to the whole regulation issue, which I think is going to be really fundamental with all these things out there. But if there's going to be a lot more regulation, that means less opportunity for growth. And obviously Donald Trump doesn't want to see that happen. Yes, in some way over-regulation will kill the market. This is the thing that I also agree with. I mean, there should be a balance. We should strike a balance between promoting financial innovation, but also making sure that the financial innovation has value and it is secure for everyone.

[00:32:21] And it's ethical and all the rest of it. But that's an enormously difficult thing to do. Oh, don't bring ethics into it. We can't do that as well. What are you doing? It is very difficult, isn't it? But I mean, I guess. So final question there, Daniela. Do you think, you know, it's a bit overblown? I mean, we're not preparing for the dot-com bomb, are we? But do you think there's, you know, has DeepSig been a wake-up call to an industry that perhaps has been a little bit bloated, dare I say?

[00:32:49] You know, these companies have been around for a long time. They've become very big. They've found very expensive ways of doing stuff. Is that part of the danger? Yes. I mean, as I said at the start, one of my expertise is a special push for acquisition companies. So even there, but even the same thing is in the dot-com bubble. There was always a bubble because every time that investors, they put money into the future of a company, they do a bet.

[00:33:18] In reality, they are making a bet. So it's something that we don't know how it works. The important point underneath is the value of the company, the real value of the company. In the dot-com bubble, many companies, they didn't have such value. But then at the same time, we had Amazon, Google. So in reality, Amazon and Google are from dot-com bubble.

[00:33:42] And the same thing we can say for special purpose acquisition companies, we had very good projects in the U.S. Some of them still today, they became important. So there are some benefits into this. The problem is to strike this kind of balance. So to be very sure that the regulator is there, possibly even without regulation, but by enforcement.

[00:34:06] To make sure that especially retail investors are not damaged, for example. So people that they cannot have financial expertise, but also they don't have the abilities to sustain a loss. And sophisticated investors, of course, I mean, they can play with smart money. So in reality, they are able to do that. But on the other hand, I mean, it's also true that it's not good to allow without any limit their involvement.

[00:34:35] It's that balance between regulation and allowing innovation. On a global basis, allowing, yeah, and somehow allowing China in if they have a solution. But that matches the regulation. Indeed. Daniele, thank you so much for talking to us. Thank you. Walking us around this, and I'm sure it'll come back up and we'll talk to you there. It'll be the story of the year, isn't it? Well, apart from tariffs. Well, there is that. There might be a bit of a fix on that as well. Good to talk, Daniele. We'll catch you soon. Thank you. Now, I'll tell you what, Roger. Away from artificial intelligence.

[00:35:02] And this thing called, and you may have been there, the real world. Oh. Yeah. Where? It just got out the door there. It's out there. And I've seen, I've noticed in my back garden, daffodil shoots starting to come up. So there will be some colour happening now. Yes. Spring is round the corner. Yeah. Which brings us to the subject we're going to talk about next week, funnily enough, which is wildlife. That was one of our best segways. Yeah, it was. It was just there. It was wildlife. Boatry motion. Wildlife in Britain. Is it?

[00:35:27] I mean, we have here terrible things about, you know, various species disappearing or becoming hugely endangered. But there are also good stories. Things about various animals reintroduced, but also animals that seem to be doing better. So we thought, you know, a bit of an audit. Where are we? What's going on? I mean, are the birds? Are we still losing the bees? Yes, exactly. Because as I understand it, the bees go. We're all dead. Completely dead. Yes. Yes. So, you know, what is the current state of British wildlife? Are we doing it more harm than we used to? Are we actually improving matters in some areas?

[00:35:57] And where are the risk areas? And should we bring back some of the old creatures? I think bears, wolves and lynxes in your garden would be my best. Yeah, it would be fantastic. I don't know what it's going to do for my cat, but maybe I shouldn't have my cat. Well, it could become a wild cat and that would be fine. Yeah. Okay. I guess so. Yeah. Well, he's halfway there anyway, really. Although he's getting old now. A bit too old to be a wild cat. Right. We'll try and make him wild for next week on The Y-Curve. We'll see you next week. Thanks for listening. The Y-Curve.