원문 보기
Okay, over the years I have done a gazillion interviews. I've interviewed everyone except you. But I've never had more interest in what I was going to ask than I'm asking you. Everybody here has at some stage you've got to ask him this. You've got to ask him that. So let's start with what is interesting to you.
I know it's AI. I know you are and the quote is I am obsessed by realizing artificial superintelligence. Are you still obsessed? Definitely. I'm only focused on that. What does that mean? You're only focused on that. Explain. Well that would change the mankind's future forever. I want to make mankind's future better.
Happier. Fantastic future. That's what I'm interested in. So you What is the difference between AGI and ASI to those to those of us born of a different generation? Okay. Okay. What's the difference? Yeah, AGI definition is same level of as a human brain. That's a AGI, artificial general intelligence.
But people have a different point of view definition of artificial superintelligence. How super? You know, 10 times super or 100 times super? My definition of ASI is 10,000 times super smarter than human brain. That's my definition of ASI and that's coming in 2035. 10 years. 10 years from today. 10,000 times smarter.
That's my prediction. What What is different in that? What will it do that we that that we should either look forward to or be worried about? Both. Both. We should We should be looking forward to that. Of course, we have to also be careful. We have to regulate. If such a superpower comes and no regulation, it could be super dangerous.
Of course, automotive industry was very very beneficial productive to human, but we have a regulation. Same thing. No No. No, I don't I I I I With respect, because we don't have anybody in the world who's willingly building or wanting to build automobiles that are dangerous. When it comes to AI, there may be state actors, there may be independent actors that want to build AI that can be harmful to other people.
How can we guard against it? How do you guard against it? Of course, there is always some bad guy. You know, 1% 2% of human are bad guys. Okay? But 99% of human are actually good guys. Okay? So, the bad guy, of course, try to use AI for the bad intention. However, those bad guy not necessarily have enough capital.
This is a massive capital. A hundreds of billions of dollars of investment which required to make super intelligence. The bad guy, how do they sponsor? There is not enough money to sponsor the bad guys. You admitted in the past that you were not as involved in AGI frankly, because you didn't have the money.
You ran out of money. Yes. Of course, the money is not unlimited. I had to save some money for the big fight. Okay? For the big opportunity. Now I'm saving you know, tens of billions of dollars so that I can make next big move. Do you know where that is? I mean, you may not know the companies. They may not even have been started yet, but do you have a very strong compass for the direction?
Yes. Tell us. I cannot. THAT'S WHAT WE'RE ALL HERE FOR. NO, NO. I have I have my own view. Go on. No, no. ARM is our core company. Okay? ARM is has 99% or should I say honestly, almost 100% of market share of all the mobile phone chip. 90% of any, you know, high-end AI AIoT chip and all kinds of other chips.
So, ARM is going to become a AI-centric chip company very soon. So, that's one thing. Okay? But there are robotics that I'm super interested in, is AI robotics, not the traditional robotics. The artificial super intelligence comes and robotics come together, that will be a tremendous, you know, uh product.
The fascinating thing about you is you have been there, you've done it, you've had the great success, you've had the follow-on failures, and you're doing it again. No, I In between, I had a many mistakes, like WeWork and, you know, Does it annoy you? Does it annoy you when we when people focus on WeWork, which was financially big, but relatively insignificant in the great scheme of your finances?
And yet EVERYBODY SAYS, "OH, MASAYOSHI, WEWORK. OH, ALIBABA, it was lucky." Does that annoy you? Well, you know, people should have their own freedom of uh thinking. Yeah. I I learn more and more by my mistake. And I almost went bankrupt twice, uh right after the net bubble burst, uh right during the Lehman crisis.
Um I survived. I actually lost 99% of our company's value, uh but somehow I came back. Did you lose that value because you took on too much risk? Or was it foolish? Or bad luck? I don't think, you know, I made that foolish mistake in the big sense. Net bubble crashed because people's view on the value of internet company uh exceeded while the true value was still on on this process, but at the end internet was not the bubble.
Year 2000 was not bubble. After that, it's much bigger. So, my general uh direction I don't regret. Now, this is really important when we talk about AI. Because if you take, for example, Nvidia the number of people who will say, "Oh, Nvidia's in trouble. Oh, Nvidia's got pro blah blah blah blah." Oh, AI's Really, all that's happened is that the market has a overvalued too soon.
That's a good question. That's a good question. Many people will still doubt. Okay? So, I think Nvidia is undervalued. What? Undervalued. On what basis? Because the future is much bigger. But the market's concentrating, as we heard in a panel earlier, it's quarterly reports. It's Oh, even even Nvidia tells us that it's going to be longer and slower, we all get Yeah, yeah, yeah.
Nvidia's just one example. Right. Okay? What about the total uh you know gen AI future? What is the value of the future? Is this a bubble today or not? Okay? Some critics, the most negative guy says, "Oh, this AGI ASI in 10 years maybe only 5% is actually useful or, you know, replacing people's job, but 95% still the same, no change.
It takes time." So, I say, "Okay, let's accept the most negative critic's view that is only 5% change, 95% unchanged, but what is that you know, ASI investment required to take? Okay, 10,000 smarter than mankind. How many How many gigawatts? We were talking about gigawatts just before me. Yeah. I have predicted it would take 400 gigawatts of AI data centers power, 400 gigawatts.
That's bigger than total US electricity. And it will require 200 million chip. Okay? The cumulative CAPEX $9 trillion. $9 trillion. How do you recoup? How do you recoup? It's too much investment for many people's view. Yeah. I say it's still very reasonable CAPEX. $9 trillion is not too big. Maybe too small.
You like to think big. Yeah, yeah. So, let me say what is $9 trillion cumulative? But if the 5% of GDP in 10 years from today replaced by ASI what is the number? Right, 5% of global GDP is $9 trillion. $9 trillion per year. The if the ASI produce you $9 trillion per year, cumulative investment of $9 trillion, small money.
All right. Small money. You recoup in 1 year for the total cumulative. So, that means $9 trillion CAPEX investment you can depreciate 1 year revenue. That means like 50% profit. 50% profit. It is going to be 4 trillion dollar net income per year. 4 trillion dollar net income shared by new GAFA. Four companies will take 1 trillion dollar profit.
And you want to be one of them. Of course. Right. Of course. Before we finish, I just want to ask let's ask our AI friend. Here the question I posed to our AI friend is is Masayoshi Son a guru when it comes to investing. I asked this question. Can we have the answer, please? Masayoshi Son is definitely recognized as a guru of investment.
His early $20 investment in Alibaba grew to $75 billion by 2014, contributing significantly to SoftBank's success. However, for some other investments, there have been mixed results. Thank you. Wait. I've got a quiet I've got a minute left. I've got a minute left. I want to ask you after both difficult times, you went into a period of comple of reflection and contemplation.
You disappeared. You thought you went up a mountain, whatever. What did you learn about yourself? Well, I say um it's a good experience. Good experience that I get defeated. I get laughed. I get criticized. That make make think harder. That make me think I have to be stronger. And that's all good. I learn more.
I think more. And before I go sleep, I have lots of concern. When I wake up, I wake up with big smile. Ladies and gentlemen, Massa, thank you very much. Thank you very much.
Comments 0