CEOInterviews.AI
Start App
Chamath Palihapitiya
Chamath Palihapitiya
CEO, Social Capital
Watch

All-In podcast host Chamath Palihapitiya on the current state of AI

📅 Jul 14, 2026 CNBC Television 13 MIN 133455 VIEWS 35 SEGMENTS · 2 SPEAKERS
Chamath Palihapitiya, Social Capital founder and CEO, 8090 CEO and ‘All-In’ podcast host, joins 'Squawk Box' to discuss the state of the AI boom, AI capex spending, AI company valuations, and more.

What Chamath Palihapitiya said

Written from the verified transcript and checked against it. Every figure links to the moment it was said.

Chamath Palihapitiya argued that the AI ecosystem faces a pricing crisis: while a few companies compound revenue at $1 billion a day, the downstream buyers of AI tokens struggle to profit. He compared AI to oil, noting a barrel of intelligence (a million tokens) costs $26 from Anthropic or OpenAI, $56 for Anthropic's latest model, $1 from Elon Musk, $1.50 from Meta, $1 from Google, and $0.50 from Chinese providers. He predicted CEOs and CFOs are unaware of 'token maxing' and will face EPS misses from runaway opex. He said the hardware and memory complex will run for another couple of years due to scarcity, and Nasdaq will outperform the S&P. He criticized leading labs for oscillating between 'super god' and 'weapon' narratives, and praised Alex Karp for exposing the risks of relying on private company funding cycles. He also questioned the effectiveness of 'zero data retention' privacy claims.

Key takeaways

  1. A barrel of intelligence (a million tokens) costs $26 from Anthropic or OpenAI, $56 for Anthropic's latest model, $1 from Elon, $1.50 from Meta, $1 from Google, and $0.50 from Chinese providers.
  2. CEOs and CFOs likely have no idea how much token maxing is happening inside their organizations, leading to future EPS misses from runaway opex.
  3. The hardware and memory complex will run for another couple of years due to scarcity, and Nasdaq will likely overperform the S&P.
  4. Leading labs oscillate between 'super god' and 'weapon' narratives, creating ripple effects at $1 trillion market cap scale.

Numbers and commitments

FigureWhat it refers toTypeAt
$1 billion a day Revenue compounding rate for a handful of AI companies metric 0:58
$26 Price per barrel of intelligence from Anthropic or OpenAI price 1:53
$56 Price per barrel of intelligence from Anthropic's latest model price 1:53
$1 Price per barrel of intelligence from Elon Musk price 1:49
$0.50 Price per barrel of intelligence from Chinese providers price 1:53
80-95% Quality level of Google, SpaceX, and Meta models compared to leading labs metric 7:31
$2 trillion Market cap of Facebook metric 9:54
$1 trillion Market cap scale at which narrative oscillations have ripple effects metric 10:24

Chapters

  1. 0:00AI pricing and downstream profitability
  2. 3:29Token maxing and EPS misses
  3. 4:20Hardware complex outlook
  4. 5:50Model convergence and quality
  5. 7:31Leading labs' constraints
  6. 8:50Trust in lab leaders
  7. 10:24Narrative oscillations and Karp's critique
  8. 12:13Privacy and zero data retention

Questions asked in this interview

3
  1. 0:00Isn't that your beer crying your coffee?
  2. 5:34How do you see them all stacking up longer term?
  3. 8:16Do you see a shift, by the way?
Host 0:00 ↗
Go cry in your beer. Isn't that your beer crying your coffee? And we're going to get to Chamath. Thanks, Steph. Okay, okay. Thanks for bringing in our special guest this morning. He is sitting on set watching all of this play out. Hasn't been here in a bit and we're thrilled to have him back. Chamath is here. He's the founder and CEO of Social Capital, CEO of 8090, and host, of course, of the All In podcast. And Mr. Palihapitiya, you're wearing a suit, no less, for us. So I...
Chamath Palihapitiya 0:33 ↗
It's my entire job selling enterprise software. That's what I do.
Host 0:36 ↗
That's what you do. So actually, before we get into anything, and I want to get into all what's happening in the Valley, AI, everything. Going to go back in time, talk SPACs, because people have asked. I know your favorite topics, but just your take on the IBM situation, to the extent that you think it has ramifications in any larger way, or maybe it's specific to IBM.
Chamath Palihapitiya 0:58 ↗
I think the thing that we know is there are a handful of companies that are compounding revenue $1 billion a day. And at some point, the downstream ecosystem has to make money as well. And then the ultimate buyer of these tokens also has to make money. And so at some point, the question has to be asked if two or three companies are going to generate two, three, $400 billion a year, and it's doubling and tripling every year, where is the rest of this money going to be made, and who is implementing it in such a way where these profits are to be had? And I think you're starting to see a little bit of the wheels come off, just at a very basic level. If you think AI is like oil, let's just use this analogy to make it simple. A barrel of crude is what, 86 bucks right now?
Host 1:47 ↗
Right now, unfortunately, it's gone up.
Chamath Palihapitiya 1:49 ↗
Yes. No, no, WTI is 80. WTI is 81.
Host 1:53 ↗
Okay, let's say a barrel of intelligence, which is a million tokens. Okay, you can buy that barrel for 26 bucks from Anthropic, a really good model. You can buy it from OpenAI for 26 bucks. Anthropic's latest model costs you 56 bucks. Elon is selling you a barrel of intelligence for a buck. Zuck is about to sell it to you for a buck 50. Demis and Sundar are trying to sell it to you for a dollar. The Chinese will sell it to you for $0.50. So this rationalization has to happen. We have the same input that has this crazy cost. If you've made a bet very early around one of these folks that are selling extremely expensive barrels of intelligence, and you try to pass through the cost, you may run into some downstream difficulty, and that has to play itself out.
And you think that's playing itself out in the IBM story, or you're saying that's just a broader...
Chamath Palihapitiya 2:44 ↗
That's just a broader comment. I think Arvind has done a great job. I mean, look, it's undeniable the trajectory of the business. And he's been able to pivot it and orient it around cloud, right? And I think he deserves a lot of credit for that. But the reality is everybody that's in this ecosystem, me included, we're all struggling to figure out how do we price this stuff so that the ultimate buyer is making more money, is growing faster? And that is still a question mark.
Host 3:10 ↗
I think you're getting to a point that Alex Karp, who was just here about a week and a half ago, was making about the sort of token maxing conceit, which is that actually some of these companies are this is to the Anthropic OpenAI piece of it, where you have CEOs, and I've now talked to them, too, who say the math is not math for us.
Chamath Palihapitiya 3:29 ↗
The CEOs and CFOs, in my opinion, probably have no idea how much token maxing is going on inside of their organizations. I suspect what will happen is one day you're going to have a miss and EPS will be off by a few pennies, and the CEO will say to the CFO, what happened? Where did all of this incremental opex come from? And they'll trace it to the $50 barrel of intelligence versus the $1 or 50 cent barrel of intelligence? That hasn't happened yet.
Host 3:58 ↗
This is interesting. There's a software stack. What is also interesting. So it's a Dow component, IBM. So the Dow is down 500 points. Just do the math, 50 points times whatever the divisor is. The Nasdaq was up more than 100. It is still. So if it is a question mark about technology in general, you're not seeing it yet in the Nasdaq. And that big 500 point drop in the Dow is because...
Chamath Palihapitiya 4:20 ↗
I think in fairness, the Nasdaq has a lot of other stuff, which is...
Host 4:24 ↗
It's not software.
Chamath Palihapitiya 4:25 ↗
All the hardware, all the memory, all the chips, that entire complex is still going to make a ton of money. Right? The question is, again, eventually, right? If you buy oil, you need to put it into a new engine. That engine has to make you go faster, right? It has to, but it sounds like you're...
Host 4:40 ↗
It sounds like you're suggesting that the LLM complex at the high end, which basically is Anthropic and OpenAI in terms of the pricing of whatever that is, and that's still in the private markets, is going to break. Is that what you're trying to...
Chamath Palihapitiya 4:52 ↗
I think that's what you're trying to get me to suggest.
Host 4:55 ↗
No, no, no, no.

17 more exchanges in this transcript

Sign in free to read the rest of this interview. No card required.

Sign in to read the full transcript

Cite this transcript

APA, MLA, BibTeX
APA

Palihapitiya, C. (2026, July 14). All-In podcast host Chamath Palihapitiya on the current state of AI [Interview transcript]. CNBC Television. CEOInterviews.AI. https://ceointerviews.ai/interview/1085556/

MLA

Chamath Palihapitiya. "All-In podcast host Chamath Palihapitiya on the current state of AI." CNBC Television, 14 Jul. 2026. Transcript, CEOInterviews.AI, https://ceointerviews.ai/interview/1085556/.

BibTeX 
@misc{palihapitiya2026_1085556,
  author       = {Chamath Palihapitiya},
  title        = {All-In podcast host Chamath Palihapitiya on the current state of AI},
  howpublished = {Interview transcript, CNBC Television. CEOInterviews.AI},
  year         = {2026},
  month        = {jul},
  url          = {https://ceointerviews.ai/interview/1085556/},
  note         = {Speaker-attributed transcript with timestamps}
}