Late last week, in an unusual meeting before a small audiencethis editor spent an hour with Sam Altman, former president of Y Combinator and, since 2019, CEO of open AIthe famous company that he co-founded with Elon Musk and many others in 2015 to develop artificial intelligence for the “benefit of humanity”.
The crowd wanted to know more about his plans for OpenAI, which has taken the world by storm in the last six weeks due to the public release of its ChatGPT language model, a chatbot that has educators and others dazzled and alarmed. (OpenAI’s DALL-E technology, which allows users to create digital images simply by describing what they imagine, generated only slightly less excitement when it was released to the public early last year.)
Because Altman is also an active investor, whose biggest return to date has come from payments startup Stripe, he said at the event, we spent the first half of our time together focused on some of his most ambitious investments.
To learn about these, including a supersonic aircraft. company and a startup that aims to help create babies from human skin cells, you can tune in to the 20-minute video below. (You’ll also hear Altman’s thoughts on Twitter under the leadership of Elon Musk, and why Altman is “not very interested” in crypto or web3. “I love the spirit of the web3 people,” Altman said with a shrug. ” But I don’t intuitively feel why we need it.)
We’ll be featuring more of our fuller conversation soon. In the meantime, below is an excerpt from our discussion of one of Altman’s biggest bets: a nuclear fusion company called Helion Energy which, like OpenAI, aims to turn a long elusive promise, this one of abundant energy, into reality. The excerpt has been lightly edited for length and clarity.
What makes a Sam Altman deal?
I try to do the things that interest me at the moment. One of the things that I’ve realized is that all the companies that I think I’ve added a lot of value to are the ones where I like to think about in my spare time going on a walk or whatever, and then send messages of text the founders and say, ‘Hey, I’ve got an idea for you. All founders deserve an investor who thinks of them as they walk. So I’ve tried to stick to the things that I really love, which tend to be hard tech, [involving] years of R&D, [is] capital intensive or is it some kind of risky research. But if it works, it really works.
One particularly interesting investment is Helion Energy. He’s been funding this company since 2015, but when he announced a $500 million round last year, including a $375 million check from him, I think he surprised people. Not many people can write a check for $375 million.
Or, like many people who [invest it] in a risky merger venture.
What have been your most successful investments to date?
I mean, probably based on multiples, definitely based on multiples: Stripe. I also think it was my second investment, so it seemed a lot easier. This was also a time when valuations were different; it was great. But, you know, I’ve been doing this for about 17 years, so there’s been a lot of really good ones, and I’m really grateful to have been in Silicon Valley at what was such a magical time.
Helion is more than an investment for me. It’s something else besides OpenAI that I spend a lot of time on. I’m super excited about what’s going to happen there.
The Lawrence Livermore National Laboratory had a breakthrough in nuclear fusion last month. (Using an approach involving giant lasers, their scientists announced the first fusion reaction in a laboratory setting that produced more energy used to start the reaction). I wonder what you think of his approach, which is very different from Helion’s (which is building a fusion machine that is reportedly long and narrow and will use aluminum tycoons to compress fuel, then expand it to get electricity out of it).
I am super happy for them. I think it is a very good scientific result. As they said themselves, I don’t think it’s commercially relevant. And that’s what excites me: not making fusion work in a lab, although that’s great too, but building a system that works at very low cost.
If you look at past energy transitions, if you can bring costs down for one new form of energy, you can take over everything else in a couple of decades. And then also a system where we can create enough power and enough reliable power, both in terms of the machines not breaking down, not having the intermittency or the need for solar or wind energy storage or something like that. If we can create enough for Earth in, like, 10 years, and I think that’s actually the toughest challenge facing Helion as we outline what it takes operationally to do that, to replace all of the current generative capacity on Earth. with fusion and to do it really fast and to think about what it really means to build a factory that is capable of producing two of these machines a day for a decade; that’s really hard, but it’s also a super fun problem.
So I’m very happy that there is a fusion race, I think it’s great. I am also very happy with solar power and batteries are getting so cheap. But I think what will matter is who can deliver cheaper and sufficient energy.
Why is Helion’s approach superior to what dozens of nations are working on in the south of France?
Yeah, well, that thing, iter, I think it will probably work, but from what I was saying before, I think it will be commercially irrelevant. They too [themselves] I think it will be commercially irrelevant.
What excites me so much about the Helion is that it is a simple machine at an affordable cost and a reasonable size. There are a lot of different elements apart from the giant. [experimental machine being developed by these nations], but one that is very cool is that what comes out of the reaction is charged particles, not heat. almost everyone else [alternatives], like a coal or natural gas plant or whatever, produces heat that drives a steam turbine. Helion produces charged particles that push the magnet and conduct an electric current through a wire. There is no heat cycle at all. And so it can be a much simpler system, much more efficient.
I think you’ve missed out on all the merger discussion, but [is] really great. It also means that we don’t have to deal with a lot of nuclear material. We never have hazardous waste or even a dangerous system. You could touch it shortly after it turns off.
is building a great facility right now. Have you already proven his thesis?
We’ll have more to share there shortly. . .
My general approach is, if there’s an area that I think is really important, like energy for example, I try to find the best fusion and fission company that I can. They are competitive in the sense that they are both trying to generate cheap power, but we desperately need more cheap power. It is a huge market; I think both can work.
