The Journal. - Sam Altman’s $7 Trillion ‘Moonshot’

Episode Date: February 14, 2024

OpenAI and its CEO Sam Altman kicked off an AI revolution with the viral ChatGPT. Now, Altman has set his sights on another ambitious goal: Raise up to $7 trillion to overhaul the world’s semiconductor chip industry. WSJ’s Keach Hagey explains what the plan entails, and why skeptics think it will be an uphill battle. Further Listening: - Artificial: The OpenAI Story  Further Reading: - Sam Altman Seeks Trillions of Dollars to Reshape Business of Chips and AI  - Raising Trillions of Dollars Might Be the Easy Part of Altman’s Chip Plan  Learn more about your ad choices. Visit megaphone.fm/adchoices

So, I'm going to say a word, and then can you tell me what you think of? Okay. Um, Java. Script? Oh, cookies. Google? Chips? Tortilla.

That's our colleague Keech Hagee. And we weren't there to talk about tortilla chips. We were there to talk about a different kind of chip, computer chips, specifically high-powered semiconductor chips. Demand for these chips has soared because of the explosion of artificial intelligence. But supply isn't keeping up.

And one person who says there needs to be more chips is Sam Altman, the CEO of OpenAI, the leader of the AI revolution. There are not enough chips to get to AGI, or Artificial General Intelligence, the stated goal of the company he leads. Artificial General Intelligence means AI can beat humans at certain tasks, and it's extremely computationally expensive to do this research. Extremely computationally expensive to do this research. And it's been set up the main choke point of this technological journey that they're on. So one way to solve it is to vastly expand the amount of chips. Now, Altman has a plan. He's working on a new project to fix the chip shortage.

But there's a hitch. This whole project might cost as much as five to seven trillion dollars. Seven trillion. That sounds like a lot of money. Indeed, it is a very great deal of money. What else is seven trillion dollars? How does that money compare to, say, the economy of a country? It is the size of the economy of some countries. It's definitely country-level scale number, not a company-level scale number. How possible is it to raise $7 trillion? I truly have no idea. It would strain the very fundamental financial underpinnings of our global economy, right?

I mean, there's no precedent for it. Welcome to The Journal, our show about money, business, and power. I'm Kate Leinbaugh. It's Wednesday, February 14th. I'm Kate Leinbaugh. It's Wednesday, February 14th. Coming up on the show, why Sam Altman is trying to raise as much as $7 trillion, yes, trillion dollars. Introducing TD Insurance for Business with customized coverage options for your business. Because at TD Insurance, we understand that your business is unique,

so your business insurance should be too. Contact a licensed TD Insurance advisor to learn more. OpenAI CEO Sam Altman kicked AI into the mainstream with the launch of ChatGPT. Before that, he spent years as the head of Y Combinator, a startup accelerator. He is a person with big vision, big dreams, and right at the center of Silicon Valley. Would you say he's a person who thinks outside of the box? Absolutely. There are many jokes that Sam can see the future because he spent years listening to startup founders pitch him. And startup is, by definition definition a bet on knowing something about

the future that other people don't know. And he's invested in, you know, dozens, hundreds of startups of all shapes, each of which has some sort of futuristic vision of one kind or another. What are some of the big futuristic projects other than AI that he's invested in? Oh, there are really too many to count. I think we've all lost sight of the appropriate ambition level here of how much of an impact we could make.

One of his personal favorites is he's invested in a startup that's trying to make cheap, repeatable nuclear fusion, which so far is not something that science can do. But if we can get fusion to work, and if we can make enough of it for the world, and if it can cut the energy cost 10x plus, that's pretty great. He has invested in a startup that is trying to extend the human lifespan by 10 years. I don't identify as a transhumanist. I think humans are really great.

by 10 years? I don't identify as a transhumanist. I think humans are really great. But I do agree that, you know, extending healthspan and augmenting humans is a very good thing to do. Right now, Altman's main focus is running OpenAI and expanding the power of artificial intelligence. But inside his company, there has been tension around resources.

One of the things which the journal has reported is that there was sort of jockeying between researchers. There was some sort of bad feeling among different teams. And over and over again, what we had heard is like the source of much of this bad feeling is fighting over compute. And by compute, you mean chips. Right. So people would say, oh, well, Sam said our team can have this compute.

And then like a few days later, like, oh, no, we can't. It has to go somewhere else. The fighting over compute we found was like a constant source of conflict and tension within the company. For OpenAI's researchers to do all the work they want to do, they need more chips. But the specific chips for AI are hard to come by. And to get more, you need more chip factories, more skilled workers, and a lot of money. So Sam has been in talks to overhaul the global microchip industry to vastly expand its capacity so that it might better serve AI and to also build other things that would better serve AI. And do you have a sense of like whether he had a pitch or like how his vision came about?

So we've heard various bits and pieces of the pitch in the course of our reporting. Broadly speaking, it involved multiple trillions of dollars. It involved a vast expansion of the world's current ability to make chips and would involve some fundamental partnerships between OpenAI and the current chip makers to basically split the cost of capital expenditure for making new chip foundries. Those foundries, the factories where chips get made, are crucial to Altman's plans. The chip shortage has been in the spotlight for a while now. Two years ago, President Biden signed the $53 billion Chips Act. The Chips and Science Act supercharged our efforts

to make semiconductors here in America, those tiny computer chips, smaller than a fingertip, that are the building blocks for a modern economy, powering everything from smartphones to dishwashers to automobiles. Biden was trying to address the chip shortage as well as national security concerns. He wants to move chip production to the U.S. from Asia, where almost all chips are made. But it's been slow. So it's been a year and they still haven't announced who is going to get the grants. There have been delays on the projects that have been established on the

ground as part of this broader initiative. So, I mean, the bottom line is it's very hard to do this in the United States of America because the salaries are higher and you don't have the trained workforce and you don't have the knowledge base. So it's like many things governmental. It's been slow and creaky and not perfect. Altman is looking to move faster, and his ambitions are huge. He's talked about building dozens of advanced chip factories, which generally cost about $10 billion to build. And the factories are just the beginning.

It's chips, it's power, it's data centers. It's kind of all the infrastructure around making AI, trying to gather a bunch of investors to do a major investment to build it out. Is it all just like cash or would it be debt? Or do you have a sense of how that works so according to our reporting in sam's thinking a lot would be debt he's gonna borrow seven trillion dollars some portion wow so this plan the seven trillion plan, to absolutely reimagine the global chip market, how likely is this plan to work? I would say it's a moonshot.

It's a long shot. So where do you go to raise money for a moonshot? That's next. your sunny days delivered with uber eats what do we mean by almost well you can't get a well-groomed lawn delivered but you can get a chicken parmesan delivered a cabana that's a no but a banana that's a yes a nice tan sorry nope but a box fan happily yes a day of sunshine no a box of fine wines yes uber eats can definitely get you that get almost almost anything delivered with uber eats order now alcohol and select markets product availability may vary by Regency app for details. This ad was expressly recorded to create a sense of simplicity. Just a few simple sounds.

No complexity. Like Neutral. No complexity. Hmm. Like neutral. Made with just vodka, soda, and natural flavor. Neutral. Refreshingly simple. OpenAI has raised a lot of money, like a $13 billion investment from Microsoft.

But it doesn't have $7 trillion. So Altman had to find some people to help pay for his chip dream. In January, he toured chip factories in South Korea and met with people in Taiwan and the Middle East. One major potential investor is the government of the United Arab Emirates. So among the folks he approached is Sheikh Tanoun bin Zayed al-Nayan, who is a very powerful figure in the UAE. a very powerful figure in the UAE. TBZ, as he is known, he is the brother of the current ruler of Abu Dhabi.

And he sits atop this large financial empire that has tentacles and all kinds of different businesses in the UAE, including this company G42, which is like an AI tech company. And according to our reporting, you know, Sam has met with him, a very powerful figure, in his talks to try to get more investment for chips. Altman has pitched the idea to chip makers around the world, including some of the biggest, like Taiwan's TSMC. He's proposing that chip makers drastically increase their production, and OpenAI would agree to be a significant buyer of those chips.

Has OpenAI or Altman said anything about this plan? OpenAI or Altman said anything about this plan? So they actually gave their first public on the record statement to the journal about this, just acknowledging that the talks were happening at all. Last week was the first time that OpenAI came forward and said, yes, we are having productive conversations with folks to, as I said, strengthen the infrastructure around AI. And they were careful to say, and we will keep the U.S. government apprised of any and all developments.

So let's talk about the reaction to this plan. So it's been kind of fascinating, actually, the reaction that I've seen. We have heard from the chips industry that, wow, you know, there are so many things that go into making chips beyond money. There is expertise, is years of planning and recruiting very, very highly skilled talent. And even if you had all the money in the world, it's not going to buy you the number of people that are actually qualified to do the most difficult manufacturing process on planet Earth. So there has been skepticism, I would say,

from the existing chips industry. Recently, Jensen Huang, the CEO of chipmaker NVIDIA, weighed in. His company plays a big role in the world of chips and AI. They are the makers of these GPUs, which are these special kind of chips that are really excellent at running these very complicated models, large language models. They're the main makers of it, so they hold the keys. Huang was asked about Altman's chip initiative earlier this week during a summit in Dubai. He said the plan, with its $7 trillion price tag, was inflated because making chips will get easier as the technology advances.

If you just assume, you know, that computers never get any faster, you might come to the conclusion, assume, you know, that computers never get any faster, you might come to the conclusion we need 14 different planets and three different galaxies and, you know, four more suns to fuel all this. But obviously, computer architecture continues to advance. There's another potential hurdle for Altman's plan. If the UAE provides funding, that could raise red flags with the U.S. government. One of the issues is that the U.S. government is very wary of who controls the chip industry globally, and particularly about Middle East with their ties to China,

and especially about some of the folks that Sam has been talking to in the UAE. So it's a big question whether the U.S. government will, how it will feel about a government like the UAE being a significant investor in an initiative like this. being a significant investor in an initiative like this. G42, the Emirati company that OpenAI is working with, is already facing some scrutiny in Congress. Representative Mike Gallagher in Congress recently wrote a letter to the Commerce Department asking for the government to investigate G42, this Abu Dhabi-based tech company, sort of alleging that

it had too many ties to China and maybe needed to have trade restrictions against it. So I'm not saying that this is happening, right? But these questions are being raised and there are concerns throughout the government about the closeness of the UAE and the Chinese tech sectors and a feeling that if we tie up with the UAE, that they're just going to pass this tech right to China. G42 declined to comment on the CHIPS initiative and didn't respond to a request for comment on Gallagher's letter. Over the weekend, Altman pushed back against criticism of his plan. He tweeted, You can grind to help

secure our collective future or you can write sub stacks about why we are going to fail. Okay. So he's sort of, what's he doing? Throwing shade on the haters? Yeah, pretty much, right? I assume that this is a subtle retort to them. Now, you've covered business and tech for some time now, Keech. Have you ever seen anything like this? I have not. What goes through your mind? I think Sam has had a lot of success in constantly moving the goalposts sort of beyond where people think they could ever be. I was going back this weekend and looking at this famous blog post he wrote, where he basically said, a good thing to do in life is to take whatever goal you have and add a zero to it. Whatever you think you can do, 10x. 10x that. 100x it.

This is the Silicon Valley way during a long period when interest rates were near zero and the stock market was going crazy and success. We got success. So somebody came up to him and was like, well, it would cost $700 billion to really change things in the chip industry. And he was like, sweet, add a zero, I'm in for $7 trillion. Maybe. Maybe. Seems kind of plausible. What do you think all this says about this moment

and about Sam Altman and his power and influence? this moment and about Sam Altman and his power and influence. The fact that the business world is discussing this as a seriously plausible plan, given that numbers relationship to like the amount of money that's actually even circulating in the world um just shows how powerful sam is and how much cred he has basically gained from pushing out chat gpt and kicking off this generative ai revolution that was something that many people who had been following tech for decades thought was not possible, right? There are things that chat GPT can do that people who have been really closely following this forever

were shocked by. And once you're shocked by something like that once, I feel like it sort of makes, it opens you up, you know? Yeah. To more possibilities, right? What you thought was impossible suddenly became possible. So it makes you start asking, well, what else that I thought was impossible might be possible?

Sam's career is sort of like a very public act of doing that, of asking that question over and over again. That's all for today, Wednesday, February 14th. The Journal is a co-production of Spotify and The Wall Street Journal. Additional reporting in this episode by Asa Fitch. Thanks for listening. See you tomorrow.