Gambling Man: Masayoshi Son’s Bold AI Vision For 2025 & Beyond | Lionel Barber
48:06 Share

The ECB is ready to do whatever it takes to preserve the euro. And believe me, it will be enough. Thank you. Let's close this door. Very pleased to be joined by Lionel Barber, former editor of the Financial Times and author of Gambling Man, the Secret Story of the World's Greatest Disruptor, Masayoshi Son. Lionel, great to have you on Monetary Matters. Welcome. Thanks, Jack. Great to be here.

Tell us about your book about this financier, Masa. He's not as well known in the West relative to his impact. Why is he so significant in the history of financial markets, not just in Japan, but around the world? Think of him in terms of a Forrest Gump-like character who's always there or thereabouts at the heart of the action in all the big technological shifts and shaping of the last four decades.

He's born in Japan. He's actually Korean Japanese. So he's an outsider, spent six years in America, Berkeley, California. So he speaks fluent English and then becomes this tech middleman and investor. So he starts in software distribution, but then he gets into dotcom stocks.

makes billions there, is the richest man in the world for three days in February 2000, then blows it, loses 97% of his paper wealth, then comes back, he's in telecoms, then he gets into global tech investing around AI, creates the Vision Fund, blows it again, and now he's back. There you saw him just the other day with President Trump

first in Mar-a-Lago and then in the White House, and he's promising to invest $500 billion in AI infrastructure. That's a sort of snapshot of Masayoshi Son gambling man.

And he is a gambler, isn't he? Just talk about some of the excessive, very large amount of risks that he's taken across his career. And might those exceed the risk limits of other legendary investors? Well, he's always been a big borrower. And remember that Japan had record interest rate, low interest rates for a long time. And he played the currency market so he could buy American assets like a chunk of Yahoo or

He bought into Alibaba, $20 million, then $80 million, turned into $130 billion once Alibaba went public. He had about a third to 40% or so of Alibaba. He has also had some bets that have gone spectacularly wrong. Jack, you may have heard of WeWork. He ended up losing probably about $14 billion on Adam Neumann's WeWork.

But I think overall, you have to say that he's taken some big gambles. My favorite story, by the way, is back in 1996 when he's got a chunk of Yahoo and he wants to buy more of the stock. And he says to the co founder, I'll give you $100 million. And Jerry Yang, who I interviewed for Gambling Man, says, well, we don't need $100 million. And Masa looks at him and says,

Everybody needs $100 million, which is probably true. And then threatens to put the $100 million in his rival. That's often a technique he uses. So he is a big gambler, but he's also capable. And I think that's what makes him a remarkable entrepreneur of actually operating serious businesses, like in telecoms, like in mobile internet. And from that story, I think he said, if you don't invest, I'll invest in your competitor and I'll kill you. Yeah.

Yes, that's a very good and accurate quote of Marcel. And I believe that in his 40-year career as an entrepreneur and investor, he's used that threat several times. My sense of him, Lionel, before reading the book, he's very...

smiling and always very happy presenting so I was surprised that he's he's got that layer to him but reading your book it didn't see it doesn't seem like he he had takes that menacing tone very often it's it's he he saves that card for to play it only sparingly right yeah he likes to play Mr nice guy he likes to play the jester the Joker I mean he speaks fluent English and he he uses

that joking role to get people to underestimate him. I mean, let's not forget that he is a student of Sun Chi's work, The Art of War. He likes to get people to underestimate him. But when it comes to a hard-boiled negotiation, when he's got money at stake,

and i heard this from for example joe sai the co-founder of alibaba who i interviewed you know he can seriously play hardball and if you as i talked to mike moritz several times the um legendary investor at sequoia i mean mike has also been in the in the hot house negotiation with him and seen that very steely side he won't he won't go short changing on his short and his self-interest

So he negotiates very toughly. But is it fair to say, Lionel, that what he doesn't negotiate on as much as maybe other investors is a very important thing of valuation. He wants to get into the deal. He wants to make the deal as big as possible. But often he doesn't care about valuation. And there are numerous examples throughout his career, aren't there, of him overpaying at least nominally for deals or not negotiating at all. Yeah, he loves to come out with a huge number of

which the founder is literally gobsmacked, can't believe that he's offering this money. I mean, I talked to a very promising entrepreneur, Herman Narula of Imagine, tech entrepreneur, and he's flying to Tokyo hoping to get 50 million investment. And Masa insists that he should have 500 million.

On Arm, the great advanced semiconductor chip designer, he offers $32 billion in cash. The board is completely flummoxed because that's money beyond what they think. But he pays big money because I think it's partly for status.

Like, I've got the money. I'm the guy with the money in the room. Remember, he created the Vision Fund, you know, with a lot of it with Gulf money from Saudi Arabia and the Emiratis. He put up a near $100 billion Vision Fund. He put up just over $30 billion, but two-thirds of it at least was from the Gulf. So he does that to sort of overwhelm people.

But I also think, and this is important, that he looks at the value down the road. So what may appear to be very expensive at the time, for example, Yahoo, I mean, Yahoo puts in 100 million. Guess what? When Yahoo is 98, 99, 2000, look at the valuation. So he just has a longer term perspective and perhaps more

Also a different perspective on value. He's saying, "I'm interested in growth." What is kind of the key to his success? Because I imagine there are more than one person in the world who understands technology as an investor and is willing to take very high levels of risk. I was about to say insane levels of risk. But what is kind of Masa's secret? First of all, he can point to several spectacular bets.

like Alibaba that I mentioned. And so when people are challenging him, as they did with WeWork, he says, well, look, everybody told me I was wrong with Alibaba and it turned out to be spectacular. So that's one angle. I think also he's been a genius at leverage of financial engineering. So he takes on debt and he loads more debt on that. Then he's got something of an asset, he borrows against it.

And he has presided over some of the biggest leverage deals in history. So he bought Vodafone Japan, which was a struggling third rate telephone mobile operator. Most famous, would you believe? It was just such a rubbish sort of company. But it did have David Beckham, you know, as its chief marketing sort of.

in the marketing campaign but he paid nearly 20 billion for it which was worth more valued more than softbank but he then built a brand and that's important you know he's built the software softbank brand built a mobile phone operator from nothing to take on ntt docomo and

very importantly, persuaded Steve Jobs to give him the exclusive distribution rights in Japan of a wonder product, the iPhone. So if you put these things together, although I do call him gambling man, I think he is a real gambler and he's bet the company several times. He's also a very capable operator. And what is the cash flow or profitability profile of

a lot of the companies that he invests in. Technology, it has a reputation for at scale requiring not a lot of capital and just being basically a cash flow machine. You think of Apple or Google and the like. But a lot of those businesses, maybe, and that's because I'm too young, but I associate a telecom business as a very capital intensive, not super profitable business. And the magazine business, obviously, it has struggled a lot over the past 20 years. But in the 80s and 90s, it was growing very quickly. So were his companies

always profitable like his core companies I'm not talking about we work yeah well at the beginning uh he was he took up front a billion dollars of losses on the broadband business in Japan over four to five years I mean he had to he cut prices to get market share so

Then with Vodafone Japan, which he had a load of debt, he then securitized that against assets in the company and used the cash flow then to pay back, which he managed to do. And it's now a very successful, profitable business. Similarly, Yahoo Japan, there were some upfront losses, but it became a very successful e-commerce business.

I think with Sprint, you'd be probably even more familiar than I am with how that company was faring back in 2012, 13, 14, 15. I mean, it was loss making. It was a pretty disastrous operation. It had some very good technology called ClearNet, looking at mobile phone technology, but

The key point, again, was he did a big securitization deal to help raise money to pay back the big debt that he'd taken out to fund the purchase. And he was also then looking to merge Sprint with Deutsche Telekom.

T-Mobile's business with, sorry, to merge it with T-Mobile owned by Deutsche Telekom. Now that took six, seven years, but when he finally managed to pull it out, he had stock there. I mean, it's been a great deal for them. So overall, I think I would give him pretty high marks on the operational side.

And then another key part of being an investor is selling, knowing when to get out. Often a very difficult part. Warren Buffett has said his ideal holding time is forever, but especially when he was younger and had exceptionally high returns, he did have quite an active selling trading portfolio. What

Does Masa know how to get out? He's a believer. He doesn't care about the valuation, but wants to get in on the big deals as they grow. But as, let's say, the price can exceed the fundamental value, does he know when to get out and realize the gains? Not always. I mean, obviously, one of the worst

sales that he made. And I call this more a trade rather than a sort of investing strategy as such. I mean, he did have nearly 5% of Nvidia and he sold in 2019. And guess what that would have been worth if he'd hung on. And I think that was partly because he was doing a lot of other investing and trades and he got distracted.

even though he was told you know this is a great stock now on in Alibaba he didn't sell like Goldman Sachs Goldman Sachs sold out after sort of three or four years I think and they did pretty well something I don't know four or five X but if you look at what Massa got for his hundred million and then he obviously invested a bit more and then when he went went public

It's a phenomenal return. Now, did he sell at the peak after the I or not after the maybe I think peak was just need to check probably 2019, maybe 20 before the tech bubble burst. So he was selling down little bits, but he probably didn't time it perfectly. But you've got to give him credit for not selling.

for like nearly 20 or 15, 20 years. So I would say done some trades very well, super sell to help him buy Arm and I give him credit for not selling Arm. I mean, if you look at Arm purchase 32 billion now worth in the market, he's got 90% of it. It's worth about 150 billion, right? Not bad. Not bad at all.

Tell us about the journey of SoftBank from being an operator business, starting with...

tech software distribution and magazines and businesses that SoftBank was buying with its own capital to becoming a manager of funds, i.e. having limited partners and being a general partner in funds. And also now, how much does that, you know, how much of SoftBank are businesses it owns versus businesses it owns on behalf of other investors? Yeah. So

You've got to remember that he starts with software distribution. That's his main business. Then he starts buying some assets in America. You mentioned Ziff Davis, which was magazine publishing. That's obviously a serious business. He probably didn't see the internet coming that was going to disrupt that business. But on the other hand, it was great for finding out what was going on in the

new product lines through the advertisers. He also buys Comdex, the trade fair in Las Vegas, which was a fabulous networking opportunity. Probably, you know, he paid 800 million for it. He certainly paid more than Sheldon Adelson, or Adelson, who is the owner, expected, the casino billionaire and Republican donor and Trump friend.

But on the other hand, you'd say he got a lot of value out of that network. Now, then back in the late 90s, during the dotcom boom, he probably had investments in more than 400 companies. And some of them are crazy. They're like the dog walking app company, the food delivery business, sort of iterations of companies that would come 20 years later. Right.

Some of those are really bum assets, if I could use a technical term. So he was doing fun. He was a sort of investor at that time. And he's always been at the tables a bit. But then he really goes into operation with telecoms. And then it comes back to investing and fund management and sort of playing at the tables alongside the operational businesses around 2015-16 with the Vision Fund.

Now, I would say at the moment, and it's complicated, I would say, is it 50-50? It depends on valuations. I mean, certainly, maybe it's not 50-50, it's less than 50-50. But the operational businesses are very important. By the way, they're the security, if you like, for the lending.

I mean, if you're looking at SoftBank Mobile, I mean, this is a serious business and it was floated partly in the late teens. Very successful. Now he's got investments in about 550 companies. There are all sorts of companies. There are robotics companies, there are AI infrastructure, there are anything with an AI moniker. They've got some food delivery. They've got Graphcore in Britain. I mean,

a lot of different things. And he could monetize those things. He could try and bring some of them to IPO and some of them he'll just keep because they link into this. And we should probably talk about this. What his next vision is, how the SoftBank evolution is going, because we've talked about operational versus investment. I think now what we're talking about is SoftBank evolving into a vertically integrated

powerhouse, which encompasses energy, it encompasses data centers like AI infrastructure, robotics companies, and maybe also likely the manufacturing of superchips like NVIDIA. So that's going to be that's the new new SoftBank.

And is that underway right now? Tell us about the new Nusoft Bank. Yeah, it really is. And by the way, I can reveal exclusively on Money Matters that far from Elon Musk thinking that this is a sort of scam, he hasn't got the money or that, you know, it's all made up on the spot. Actually, Masa has been plotting this for at least nearly, I'd say, two years.

Because I did my first interview with Masa in February 2023. And I've been to Tokyo once. He was too busy to see me. It's a long way from London to Tokyo, by the way. I was hearing from his people that he had some grand plan and it was the biggest comeback coming. And I think this has gradually rolled out, this idea of, you know, I'm going to be the AI champion and I'm going to get into bed with Sam Altman.

I've been playing with chat GPT. We're about to go into a whole new age, AI arms race. We're going to move to artificial general intelligence when robots are going to be as smart as humans, if not smarter. And I'm going to be at the center of this revolution. That's what you're seeing at the moment. That's what got rolled out when Masa, Masayoshi Son,

You know, amazingly, and it was courtesy Howard Lutnick, by the way, the Commerce Secretary, ex-Cantor Fitzgerald, who got him in to see that special access with Trump, with the president at the White House. And they're offering, they're promising $500 billion investment in AI infrastructure in America over the next five years. And Lionel, like many, I saw that deal, Masa, to invest

hundreds of billions in the US, I was a little bit cynical and I thought, oh, that just means he will buy companies worth $100 billion. That doesn't necessarily mean that all across America, America economically is going to see the benefit of hundreds of billions of dollars. But are you saying that they're actually investing in operations? It's not just buying stocks or even companies? No, I think you're making an extremely important point.

that this will cover investments as opposed to just, for example, that's easy thing to understand, construction jobs, building a giant data center deep in the heart of Texas, for example. There will be jobs like that. There will be data scientists recruited,

But these gigantic sums are going to be chunks of money invested in other companies, which are part of what Masa likes to describe as AI infrastructure. Now, just for ordinary readers start here, just, you know, what does he really mean by that? Well, the analogy he draws is with the Internet.

that in the late 90s, people were saying, well, do we need all this broadband capacity and broadband? Really? The highway? Well, the answer is, I think we did because the internet exploded, mobile internet exploded. And he says the demand for energy and for

data centers that can absorb all this training for LLMs, etc. That's going to exist and that is what counts as infrastructure. I can just read it on your face. What you're about to say is, oh, that's very well, Lionel Barber, but you've bought all this hook, line and sinker. And what about China? What about Deep Sea? Because they came up with a new model which doesn't need all the energy and is sort of a lot more efficient without the advanced chips.

Well, that is part of the picture. And that does challenge a number of the sort of preconceptions that we've had about the next stage of AI development. When you ask Masa, what he says, it's going to be fine because the market will grow and will accommodate everybody. Right. You read skepticism on my face. Right. It wasn't just about China. It was the analogy of it's 1998, 1999,

telecom is needed to build out the internet. I think Masa and you are right that ultimately there was a need for that internet, but the investors in equity and debt into those companies, telecom companies, it's my understanding that they enjoyed a horrible return. So I presume that Masa has a much more sanguine outlook for the data centers because if the analogy is one-to-one, it is the users of the rails of telecom

artificial intelligence will be enormously profitable and trillions of dollars in market cap, but the construction of the rails will, like the telecom giants that lost a ton of money in the 2000s, lose money. But I presume that MASA has a much better outlook. Well, I think it needs to be pinned down on that distinction. And as you rightly say, the history of

doesn't always repeat itself, but it sometimes rhymes. And if you look at the story of the railways in America in the mid to late 19th century, those investors in the infrastructure also didn't have very good returns. And we all know what Warren Buffett has said about airlines.

So you may be right about this. And is MASA that $100 billion? And correct me if I'm wrong, Donald Trump, I believe before he became president, but maybe after he was elected, said he asked for $100 billion and then MASA brought $200 billion. I remember there was a press conference where he said, I brought double what you asked. And then is that investment, is that coming from venture capital funds, LP funds, or is it SoftBank's own money that it's going to

borrow or get by issuing shares? Well, first of all, the meeting took place before Donald Trump was inaugurated as president. This took place at his winter retreat in Mar-a-Lago. On the Sunday, very important fact, because this is where all the deals are done, Massa plays six hours of golf with the president. And then

The next day they hold a press conference. Massa promises 100 billion and Donald Trump tries to bounce him into saying 200 billion. And Massa says, well, I will do my best. Well, a few weeks later, he's suddenly promising 500 billion. So it's sort of I'll see you and raise you. I do think it's partly leveraged around

the OpenAI and Sam Altman's plans to create his own trillion dollar company. And Masa's promising to put between 15 and 25 billion dollars into OpenAI as part of the fundraising round, right? Masa's also promising a lot of money into OpenAI too. So he's sort of

It's two horses. Now, where is he going to get that money from? The answer is he's got around 40 billion on his 35, maybe 40 billion on his balance sheet. Obviously, he wouldn't want to spend all of that. He does have that. He's also got assets. He's got stock in the T-Mobile spread out of the T-Mobile Sprint deal. He's got SoftBank Mobile. He could sell some shares in that. I mean, he can raise money.

He's also, I'm told, but we haven't seen the color of that money yet, got some institutional money, perhaps Japanese pension funds. I'm not sure about Wall Street. Certainly Gulf money, whether it's PIF, the Public Investment Fund, Sovereign Wealth Fund, maybe not. But he does have Gulf money. So I think what he's trying to do, Jack,

in short, is create a sense of momentum. And he's getting a bit of pixie dust, stardust from Sam Altman. And as we know, Sam Altman, he's talking about open AI, which let's not forget, is not actually making a profit. He says it's worth $300 billion. Yes. So I think there's going to be some leveraging and some playing off

the valuation of OpenAI and other assets. And in the 90s, Masa took advantage of the very low interest rates in Japan and thereafter. Is the Japanese banking system

Just suited to lend to Masa in a certain way. Obviously, you had a huge bubble in the late 80s that popped. And then there was a many, many long year recession. But I wonder, is it in that recession? Was there just a lack of borrowers? And that's why Masa was able to borrow so much money. Why has he been able to borrow so much money? And then you talk about the details about how he's borrowed in the bank credit market as well as in the bond market.

Well, remember, this is not really commercial lending as such, direct commercial lending. What he's doing is raising money in the debt market. So he's corporate bonds. And obviously, then the banks are looking at the rating and they also expect to be repaid on a certain schedule, right? If they've subscribed to the bonds, too.

Now, what Masa did, and I described this in the book, Jack, and it's a story which is not widely appreciated in America, is the degree to which he and he hired this really tough guy, real tough guy, Yoshitiko Kitao, no longer works with him. But they basically disrupted the whole bond market in Japan.

and redress the balance or completely change the balance between borrower and bank. Because previously, you basically, you know, the banks arranged the bond deal and you had to sort of like it or lump it. You took your just desserts. Whereas what he, they managed to use some of the change in the law ushered in by a reformist party

Japanese government is trying to compete with America and Wall Street to basically disrupt the bond market and change that balance of power. And as a result, they could sort of pick their banks and they've had a good record of repaying debt. OK, they're very careful. So even when he's gone very close to the edge, they've not said they've not stiffed the banks. They've paid back. Now, you can also argue that

that maybe SoftBank is too big to fail. And the biggest bank, which is Mizzou, they've right in there. They've got a huge amount invested with, they're exposed to Masa and SoftBank. I would argue, and I talked to them about that, that essentially it's a long-term relationship.

They've been very good. And they also see SoftBank, even though the corporate hierarchy in Japan are a bit mistrustful of Masa. He doesn't make things, doesn't run a steel plant or build automobiles. He's a financial engineer. He's an investor. They kind of think he's a necessary, if not evil, he's necessary for the system. He provides some grease. He's a risk taker in a very risk averse country.

So for all those reasons, that's how you kind of explain the relationship between Masa and the commercial banks. One other point, because the banks really were underwater in the 90s after this collapse of the real estate market.

And Lionel, there's a macro story there. When reading the book in the 1990s, just how low the interest rates were. If you were to write the story through the lens of macroeconomics, there's a low interest rate in phenomena. What about the currency market, as well as the relationships between Japan and the rest of the world? I mean, even the magazine, the fact of selling software, is it...

Is that related to maybe Japan's pretty toughness on trade and how there always has to be a Japanese company that is importing? Well, let's break that down. First of all, I happen to be in Washington as the correspondent for the Financial Times in the mid 80s and through to 92.

So I saw this period straight up when people were absolutely paranoid about the Japanese. They thought they were going to take over the semiconductor business. They were worried about Japan buying huge assets like the Rockefeller Center, like Sony buying Columbia Pictures. What was happening then was, first of all, a stock market bubble. So you could raise loads of money. You could use stock for deals.

But you're also you had a very strong yen. So they and the Americans, thanks to the Plaza Accord, Jim Baker was Treasury Secretary at the time, they engineered essentially a devaluation of the dollar and appreciation of the yen. So that really killed or really hurt Japan's export business. But for someone like Masa,

who was the middleman with American companies wanting to get exposure to the mass market in Japan for say, video games or consumer software, business software. He could also use that strong yen to buy some really quite good assets in America. So that's what you see in that period. Then with the kind of semi comatose Japanese economy,

He exploits record interest, low interest rates to raise money in the markets and does very well. He also can build his broadband business on the back of that. I think later on it become you've still got the low interest rates, which helps him, helps him a lot during the global financial crisis, by the way, where he got into serious trouble.

I think then, I mean, obviously currency does fluctuate a bit. You can go from, I don't know, 85, 90 yen to the dollar to 150 recently. So that's quite a big spread. Now, interest rates have begun to increase very gently. The Bank of Japan is trying to exit this unconventional monetary policy that it's been running for more than 20 years.

And so that affects them a bit. But look, we're not talking about 3% interest rates or 5% interest rates in Japan. It's still pretty damn low. Yes. The 10-year I'm looking now is 1.44%. That is pretty damn low, as you say, but it is a lot higher than zero. And there have been very volatile moves in the currency. So is the weaker yen, has that...

How does that affect Masa's empire? And, you know, I mean, you're reading articles in Bloomberg or the Financial Times about just how, you know, sushi in New York is so much more expensive than the best sushi in Tokyo. Well, indeed, I can bear witness to that. I think New York prices are absolutely bonkers. But there we are on everything. Never mind sushi. It's... Look, he does run a sort of semi-global business. You know, his...

businesses in America of dollar cash businesses, right? I mean, he's not just looking, sitting in Japan with a weak yen or a weaker yen. So I think that gives him some protection. They've got currency hedges as well. So I don't think now with this global footprint, it's such a problem.

He's got energy business, solar business in America. Lionel, let's go back to Masa's new vision for SoftBank in 2025 and beyond. You talked about a vertically integrated AI empire, you know, in school learning about vertical integration in oil. That's okay. You own the upstream assets, the exploration and production. You own the

refining, you own the sales of it. That's what vertically integrated is. So I imagine in AI that looks like you own the energy, you own the data centers, you own the large-legged models that run on those clouds. What is his vision? And yeah, what do you know about his grand vision that our audience would be interested in and that not many people know? One of the clues to understanding how Masa sees businesses, first of all, total accessible market.

So he's always interested in the growth aspect as opposed to making money. And he's prepared to take a longer term view. Second is cover every chip on every square on the board. So he's quite happy to have, as he said, 500 companies that he's invested in. And, you know, if a couple are like Alibaba,

or Nvidia, that's okay. He'll have 98% duds. So that's a sort of moonshot attitude. But then the third is, he's always interested in spotting bottlenecks. So here, he would say, well, you know, we've got a problem that if you've got to generate, you need all this computing power, then clearly energy is a big issue.

So I'm in that business. I've been in that business ever since. And I described this in the book in Gambling Man, the seminal experience of the Fukushima earthquake when this nuclear plant basically just blows up. Masa goes massively into alternative energy and solar energy. He sold part of that business, but he's kept some. So he sees energy as a very important aspect.

But then final piece in the block is everything hooks together, both energy, computing power data center, the companies that have these large language models that need training. And then I've got the robotics company that's part of this puzzle. Now, how it perfectly fits together and slots into place.

We haven't seen that. One or two people I've talked to say, yeah, the vision sounds great, but what about the execution? Can he put it all together? All I would say this time is that this is a man who pays extraordinary attention to detail. This is a man who literally drew up sheet after sheet of

of minute designs for the new Indonesian capital of Nusantara. He has an amazing capacity to absorb detail. So I think that he has thought through a lot of the integrated, vertically integrated model. What I don't know for sure is what about this force that could come out from nowhere, this kind of

left field, deep seek phenomenon? Have they really taken that on board? Or is he so keen

to get into the AI revolution that he's been talking about for 25 years, by the way, that he may have kind of gone all in and fallen in love with Sam Altman a bit like he fell in love with Adam Neumann. And do you think that, or what are you hearing on whether artificial intelligence and building the next generation of large language models

is going to require the trillions of dollars of investment that Sam Altman says it does. I don't know about Apple, but Google, Oracle, Amazon are investing 50, 80, 100 billion dollars a year in some instances to build out this data infrastructure. But what if DeepSeek, you only need a billion dollars, quote unquote, only a billion dollars,

And also, what are your personal opinions on did DeepSeek, are they fibbing, to use the polite word for falsehood, that it actually, they only spent such a little amount and actually they have bought a lot of NVIDIA chips that they're just not owning up to? I think it's very unlikely that DeepSeek only spent $500 million in developing this. It seems fairly clear, but for something slightly less expensive,

impressive than some of the American models. They have managed to produce something pretty damned interesting with not so much cost and crucially less refined chips. And they've done that using techniques that even I, as I'm not a scientist, don't fully grasp, but you can read about it, you know,

where the sort of allocation of task is less... The best analogy is the library. Do you go beetling around every shelf, every book? Or do you sort of take it in sections of the library and then use intuitive learning to get to your best results? And the initial evidence, and remember, this is open source.

This is not propriety like Sam Altman's OpenAI, which is really impressive. But this Chinese model has taken a lot of people back. Now, there's one other new development, which I think your listeners would be really wise to look at. President Xi of China, who's presiding over a stagnant Chinese economy, just last week

or earlier this week, brought in the Chinese tech elite and sort of gave him his imperial blessing, including Jack Ma, the co-founder of Alibaba,

who'd been banished into sort of exile for criticizing the financial regulators and the payment system, you know, because he wanted to float Alipay, hugely valuable IPO in the pipeline. And it got sunk by the Chinese Communist Party and the regulators. I think that's a real sea change.

in official attitudes. It's a realization that China needs those top people to compete against America in the AI arms race. And I think, you know, it also shows that there's a lot of firepower at China's disposal. So this story ain't over.

It's not over. And how would you characterize Masa's relationship with China? He was an investor in Alibaba, but it's no secret that Japan and China have had a fraught relationship politically, to say the least. And with the...

at least rumored or understood to be closed markets of China over the past three years. Does he feel like Japanese investors and Masa don't have as much access to China's markets or is that not really true? Well, he said in an interview with me in October 2023, which is the last time I sat down for an extended conversation with him for the book, Gambling Man, that

You know, tensions were such between East and West that he'd had to choose between the US and China, and he'd chosen the US. Now, if you look at the announcement with Donald Trump and the president last week, that seems pretty clear, his decision. But he does have still some exposure to Japanese, to Chinese stocks, and he does retain a

a deep friendship and partnership with Jack Ma. So he sheltered Jack Ma, who went to Tokyo. He helped him after he got sort of essentially banished or from public view. So

I would say knowing Masa that, yeah, he's telling the truth about him making a strategic decision, but he's always hedging his options a bit. And just when you said earlier, I can break for the first time on monetary matters. What you're revealing is that Masa's son has been revealing to invest a huge amount in building this vertically integrated AI empire for at least two years, I believe you said. But also, can you reveal...

But is the cash lined up? What are you hearing? You said, OK, $40 billion in assets. Obviously, he's got a large access to the bond market and to borrowing. But, you know, is the money forthcoming? What do you think? What have you heard? So what I'm hearing is that he has some of the money, but not all of the money.

more money than maybe Elon Musk thinks he has. And Elon Musk has, by the way, got a direct interest here because he's got his own AI company and he also doesn't have best relations with Sam Altman. I don't think that Masa has lined up all these investors over the last two years. I think he's had a very tight circle of people

that he's been plotting with about how to carry off this next stage of the AI revolution. And crucially, he's got the Trump administration on side. And by the way, that does help market valuations. If you look at Twitter valuation and how much Elon Musk's personal wealth has gone up,

I mean, it's the Trump touch, so to speak, at least in the short term. So I think he then talks big numbers, then he goes up lining up the money. People say to me that they can find a way of outlining how he gets to 100 billion. Whether he gets to 500 billion inside the next four years, I think we need to take that with a little bit of salt.

Lionel, it's been such a pleasure getting a chance to speak with you. The book is Gambling Man, The Secret Story of the World's Greatest Disruptor, Masayoshi Son. I really enjoyed reading it and I think my audience will as well. So we'll include a link in the description. People should check it out. You can be found on Twitter at Lionel Barber. Thanks again, Lionel. Jack, it's a great pleasure to be on the program. I'm really pleased that you like the book. Thanks. Thanks.

Thank you. Just close this f***ing door.