Aaron Levie

And I don't think it's beyond the realm of possibility that we could be at 5 or 10 gigawatt data centers in the next, I don't know, several years. So my point is just, look, the algorithms, the chips, and the data centers are all improving or scaling at a rate of, I don't know, three to four X a year. That's 10X every two years, okay?

And I don't think it's beyond the realm of possibility that we could be at 5 or 10 gigawatt data centers in the next, I don't know, several years. So my point is just, look, the algorithms, the chips, and the data centers are all improving or scaling at a rate of, I don't know, three to four X a year. That's 10X every two years, okay?

Where people don't understand exponential progress is that if you're getting better at 10X every two years, that doesn't mean you'll be at 20X in four years. It means you'll be at 100X, 100X. So the models, the chips, and the data centers will all be 100 times more powerful in four years let's say at the end of this presidential term.

Where people don't understand exponential progress is that if you're getting better at 10X every two years, that doesn't mean you'll be at 20X in four years. It means you'll be at 100X, 100X. So the models, the chips, and the data centers will all be 100 times more powerful in four years let's say at the end of this presidential term.

So you multiply those things together, the algorithms, the chips, and then the raw compute that's available, you're talking about a million X increase, some of which will be captured in price reductions, some of it will be in the performance ceiling, and then some of it will just be in the overall amount of AI compute that's available to the economy.

So you multiply those things together, the algorithms, the chips, and then the raw compute that's available, you're talking about a million X increase, some of which will be captured in price reductions, some of it will be in the performance ceiling, and then some of it will just be in the overall amount of AI compute that's available to the economy.

But the impact of this thing is going to be absolutely massive. And I think people still don't even appreciate that fact because they don't understand exponential progress. Yeah.

But the impact of this thing is going to be absolutely massive. And I think people still don't even appreciate that fact because they don't understand exponential progress. Yeah.

What's happening right now, the reason why the progress is so rapid in coding assistance. And I think, you know, you're right that Anthropic with, was it Claude 3.7? Yeah, I think they're the leader.

What's happening right now, the reason why the progress is so rapid in coding assistance. And I think, you know, you're right that Anthropic with, was it Claude 3.7? Yeah, I think they're the leader.

And in fact, I think the Manus demo that we showed, it's not entirely a wrapper on Claude because they actually, they do a number of different things, but I think they are significantly using Anthropic for the

And in fact, I think the Manus demo that we showed, it's not entirely a wrapper on Claude because they actually, they do a number of different things, but I think they are significantly using Anthropic for the

code assistant part of it in any event the reason why the progress is so rapid with coding is because code compiles and you can determine objectively whether it works or not you can validate it and so that makes it a perfect area for ai to get better at through reinforcement learning and test time compute is ai tries a bunch of things it sees what works it sees what compiles sees what the user then accepts and then to be able to learn to iterate based on that that's why

code assistant part of it in any event the reason why the progress is so rapid with coding is because code compiles and you can determine objectively whether it works or not you can validate it and so that makes it a perfect area for ai to get better at through reinforcement learning and test time compute is ai tries a bunch of things it sees what works it sees what compiles sees what the user then accepts and then to be able to learn to iterate based on that that's why

Coding right now is really the big breakthrough application and use case. But it's not going to be the only one. I mean, math is another good area where I think AI is improving rapidly, again, because math, you have proofs, and you can look at the results and see if it validates.

Coding right now is really the big breakthrough application and use case. But it's not going to be the only one. I mean, math is another good area where I think AI is improving rapidly, again, because math, you have proofs, and you can look at the results and see if it validates.

Now, I think one of the big questions in terms of AI progress is how extensible is the progress to other areas that don't easily validate that way? So for example, legal work, is I think a really good area for AI, but how do you validate that is correct?

Now, I think one of the big questions in terms of AI progress is how extensible is the progress to other areas that don't easily validate that way? So for example, legal work, is I think a really good area for AI, but how do you validate that is correct?

Or you could hire like a thousand lawyers or experts in the area to basically do, you know, reinforcement learning through human feedback.

Or you could hire like a thousand lawyers or experts in the area to basically do, you know, reinforcement learning through human feedback.