NVIDIA Corporation

Sacks oppose ×2 ▲ on 🤖 AI compute capex supercycle has 12-24 months of runway E296 · 2026-11-07 ▶ 1:22:53

But I think by 2028, I would expect that the pipeline will either be closed or significantly reduced and that will have a big economic impact.

Jason support ×3 ▲ on 🤖 AI compute capex supercycle has 12-24 months of runway E296 · 2026-11-07 ▶ 1:12:48

So I think there's 500 to 1,000 times the need for this. When that happens, this compute is going to go wherever it needs to go.

Sacks reversal ×3 ▲ on 🏛️ The AI doomer campaign fails — no FDAI, open source survives E296 · 2026-11-07 ▶ 1:07:42

I believe that Democrats want to create a surveillance and censorship state that's turbocharged by super intelligence. I mean, that's ultimately the end state. How do you get there? You create a department, a federal department of AI. You require that all the so-called frontier models have to be reviewed and approved by that.

Jason support ×2 ▲ on 🏛️ The AI doomer campaign fails — no FDAI, open source survives E296 · 2026-11-07 ▶ 1:06:11

I think this is like you're getting into Dr. Doom scenarios. I think you're completely wrong. I think it's like a 1% chance of something crazy like this happens.

Friedberg oppose ×2 ▲ on 🏛️ The AI doomer campaign fails — no FDAI, open source survives E296 · 2026-11-07 ▶ 41:06

I still think, I could be wrong, but I still think we end up in a place where the government says we want to regulate compute.

Sacks support ×2 ▲ on 🏛️ The AI doomer campaign fails — no FDAI, open source survives E295 · 2026-10-30 ▶ 1:00:16

And I think actually that this IPO will be very helpful and clarifying on that because, again, once they're seen as a big tech monopoly or part of a huge duopoly, the arguments for giving them regulatory modes- This kind of advantage will go away.

Sacks support ×3 ▲ on 🤖 AI compute capex supercycle has 12-24 months of runway E295 · 2026-10-30 ▶ 56:19

And the net result is that the total demand just keeps going up. And the total revenue and market size keeps growing. Also, just to reiterate a point that you made earlier in the program, the new applications that are coming all consume a lot more tokens than the ones that we started with, right? Like agents use a lot more tokens than prompting.

Friedberg oppose ×2 ▲ on 🏛️ The AI doomer campaign fails — no FDAI, open source survives E295 · 2026-10-30 ▶ 36:48

I think that you're going to end up seeing compute get regulated in the future, where compute capacity is going to be assigned by government, and they're going to pass laws to try and do that post-production, that is, like, meaning they're not going to control the chips. They're going to control the actual compute layer, and they're going to determine who's going to get access to what amount of compute.

Chamath support ×2 ▲ on 🤖 AI compute capex supercycle has 12-24 months of runway E295 · 2026-10-30 ▶ 35:47

You're absolutely right, Jason. The problem is the regulatory pushback. The amount of approvals and the ability for utilities to find the staff and the ability to actually turn these things on and then the ability to get the capital equipment you need, it just isn't there right now.

Jason support ×3 ▲ on 🤖 AI compute capex supercycle has 12-24 months of runway E295 · 2026-10-30 ▶ 34:57

The token usage is 500x. That means we're going to have not enough compute, and everybody's got to build out it completely wrong. There won't be enough.

Friedberg support ×3 ▲ on 🤖 AI compute capex supercycle has 12-24 months of runway E295 · 2026-10-30 ▶ 12:53

But generally speaking, if AI is delivering value to the customers, I think this shows that there really is true productivity arising from AI, that across the whole value chain, there is such margin to be had, so much value being created, all the way through to the customers

because I think demand is insatiable. There's enough demand, there's more demand than you and I can imagine. So yes, I understand the enterprise diffusion comment, but it's going to get diffused. The question is, if every company in the world today said, I'm in, I'm ready, I'm embracing AI, let's go as fast as we can. There's not enough compute in the world to satisfy us.

Again, zero of those four have been met. We're not on track for any of those four to be met unless something is going on in the big labs that we don't know about.

Chamath support ×2 ▲ on 🏛️ The AI doomer campaign fails — no FDAI, open source survives E293 · 2026-10-17 ▶ 1:24:22

And he's the doomer. He's the flag-carrying doomer, trying to essentially have what he had the first time around, which is command and control from a global hierarchy.

Jason support ×2 ▲ on 🏛️ The AI doomer campaign fails — no FDAI, open source survives E293 · 2026-10-17 ▶ 1:11:02

They're making up scenarios that get a lot of attention. These are fantastical.

Sacks support ×3 ▲ on 🏛️ The AI doomer campaign fails — no FDAI, open source survives E293 · 2026-10-17 ▶ 1:05:38

They're trying to create a panic. This is their latest panic against open source.

Chamath support ×2 ▲ on 🤖 AI compute capex supercycle has 12-24 months of runway E293 · 2026-10-17 ▶ 44:56

I think we're capacity constrained enough through the entire pipeline.

Chamath oppose ×2 ▼ on 🤖 Nvidia risk: special-purpose silicon + Huawei E293 · 2026-10-17 ▶ 44:56

You have to think, NVIDIA has done an incredible job. Because if you look one layer beyond NVIDIA, all the packaging capacity is them, all the HBM capacity is mostly them.

Sacks support ×3 ▲ on 🤖 AI compute capex supercycle has 12-24 months of runway E293 · 2026-10-17 ▶ 21:36

So we know the demand is off the charts right now. We're seeing other indicators. So, for example, old GPUs, like even the H100s or the A100s, which are several generations old, they're still renting at premium prices right now.

I said, yeah, like, we're out of everything, you know, till 2028 He's like, oh, it's beyond 2028 Like, I don't think we get stuff back. Like, end of 2028, maybe. So everything else is sold out.

Sacks support ×2 ▲ on 🏛️ The AI doomer campaign fails — no FDAI, open source survives E292 · 2026-10-10 ▶ 1:18:55

But what do we hear about? AI is going to put everyone out of work. Total fake narrative. That was last year's hoax. Now, there's a new hoax about human extinction.

Friedberg support ×1 ▲ on 🏛️ The AI doomer campaign fails — no FDAI, open source survives E292 · 2026-10-10 ▶ 30:26

And I think that that's really why, to me, this is more bad idea than we need to kind of jump into a regulatory freak out about it.

Sacks support ×2 ▲ on 🤖 AI compute capex supercycle has 12-24 months of runway E291 · 2026-10-02 ▶ 24:30

Yeah, here's the part I agree with, is it's not just about the capability of the model, it's about how much compute you have to run it. And we want to have more compute than our adversaries.

Jason support ×1 ▲ on 🤖 AI compute capex supercycle has 12-24 months of runway E291 · 2026-10-02 ▶ 24:04

Yeah, it's an opportunity to sell into the government. Now, Google and Amazon Web Services, Crusoe, whoever, can just sell more servers.

Friedberg support ×2 ▲ on 🤖 AI compute capex supercycle has 12-24 months of runway E291 · 2026-10-02 ▶ 21:18

Therefore, we need to increase our investment in AI to maintain our cyber defense capabilities, both at the enterprise level and at the government level and potentially even at the civilian consumer level. So I think that the cyber defense necessity actually mandates that we continue to build out data centers and support it.