Nathan Lambert

So accounting profit-wise, Microsoft is making money, but they're spending a lot of CapEx, right? And that gets depreciated over years. Meta is making tons of money, but with recommendation systems, which is AI, but not with Lama, right? Lama is losing money for sure, right? Yeah. I think Anthropic and OpenAI are obviously not making money because otherwise they wouldn't be raising money, right?

So accounting profit-wise, Microsoft is making money, but they're spending a lot of CapEx, right? And that gets depreciated over years. Meta is making tons of money, but with recommendation systems, which is AI, but not with Lama, right? Lama is losing money for sure, right? Yeah. I think Anthropic and OpenAI are obviously not making money because otherwise they wouldn't be raising money, right?

So accounting profit-wise, Microsoft is making money, but they're spending a lot of CapEx, right? And that gets depreciated over years. Meta is making tons of money, but with recommendation systems, which is AI, but not with Lama, right? Lama is losing money for sure, right? Yeah. I think Anthropic and OpenAI are obviously not making money because otherwise they wouldn't be raising money, right?

They have to raise money to build more, right? Although theoretically they are making money, right? Like, you know, you spent a few hundred million dollars on GPT-4 and it's doing billions in revenue. So like, obviously it's like making money. Although they had to continue to research to get the compute efficiency wins, right?

They have to raise money to build more, right? Although theoretically they are making money, right? Like, you know, you spent a few hundred million dollars on GPT-4 and it's doing billions in revenue. So like, obviously it's like making money. Although they had to continue to research to get the compute efficiency wins, right?

They have to raise money to build more, right? Although theoretically they are making money, right? Like, you know, you spent a few hundred million dollars on GPT-4 and it's doing billions in revenue. So like, obviously it's like making money. Although they had to continue to research to get the compute efficiency wins, right?

And move down the curve to get that 1200X that has been achieved for GPT-3. Maybe we're only at a couple hundred X now, but with GPT-4 Turbo and 4.0, and there'll be another one probably cheaper than GPT-4.0 even that comes out at some point.

And move down the curve to get that 1200X that has been achieved for GPT-3. Maybe we're only at a couple hundred X now, but with GPT-4 Turbo and 4.0, and there'll be another one probably cheaper than GPT-4.0 even that comes out at some point.

And move down the curve to get that 1200X that has been achieved for GPT-3. Maybe we're only at a couple hundred X now, but with GPT-4 Turbo and 4.0, and there'll be another one probably cheaper than GPT-4.0 even that comes out at some point.

Yeah, to do things like reasoning, right? Now that that exists, they're going to scale it, they're going to do a lot of research still. I think the... People focus on the payback question, but it's really easy to just be like, well, GDP is humans and industrial capital, right? And if you can make intelligence cheap, then you can grow a lot, right? That's the sort of dumb way to explain it.

Yeah, to do things like reasoning, right? Now that that exists, they're going to scale it, they're going to do a lot of research still. I think the... People focus on the payback question, but it's really easy to just be like, well, GDP is humans and industrial capital, right? And if you can make intelligence cheap, then you can grow a lot, right? That's the sort of dumb way to explain it.

Yeah, to do things like reasoning, right? Now that that exists, they're going to scale it, they're going to do a lot of research still. I think the... People focus on the payback question, but it's really easy to just be like, well, GDP is humans and industrial capital, right? And if you can make intelligence cheap, then you can grow a lot, right? That's the sort of dumb way to explain it.

But that's sort of what basically the investment thesis is. Um, I think only Nvidia is actually making tons of money and other hardware vendors. Um, the hyperscalers are all on paper making money. Uh, but in reality, they're like spending a lot more on purchasing the GPUs, which you don't know if they're still going to make this much money on each GPU in two years. Right.

But that's sort of what basically the investment thesis is. Um, I think only Nvidia is actually making tons of money and other hardware vendors. Um, the hyperscalers are all on paper making money. Uh, but in reality, they're like spending a lot more on purchasing the GPUs, which you don't know if they're still going to make this much money on each GPU in two years. Right.

But that's sort of what basically the investment thesis is. Um, I think only Nvidia is actually making tons of money and other hardware vendors. Um, the hyperscalers are all on paper making money. Uh, but in reality, they're like spending a lot more on purchasing the GPUs, which you don't know if they're still going to make this much money on each GPU in two years. Right.

Um, you don't know if, um, All of a sudden, OpenAI goes kapoof, and now Microsoft has hundreds of thousands of GPUs they were renting to OpenAI that they paid for themselves with their investment in them that no longer have a customer. This is always a possibility. I don't believe that. I think OpenAI will keep raising money.

Um, you don't know if, um, All of a sudden, OpenAI goes kapoof, and now Microsoft has hundreds of thousands of GPUs they were renting to OpenAI that they paid for themselves with their investment in them that no longer have a customer. This is always a possibility. I don't believe that. I think OpenAI will keep raising money.

Um, you don't know if, um, All of a sudden, OpenAI goes kapoof, and now Microsoft has hundreds of thousands of GPUs they were renting to OpenAI that they paid for themselves with their investment in them that no longer have a customer. This is always a possibility. I don't believe that. I think OpenAI will keep raising money.

I think others will keep raising money because the returns from it are going to be eventually huge once we have AGI.

I think others will keep raising money because the returns from it are going to be eventually huge once we have AGI.