Video · 2026-07-11 · 1h 42m · 18 moments

OpenAI vs Anthropic IPOs, Anthropic $3T, Zuck's Price War, China Ends Open Source?, Trump Accounts

✦ AI generated

GUEST Brad Gerstner GUEST Sacks GUEST David Sacks GUEST Chamath Palihapitiya

timeline · colored by role

Anecdote◆

AI companies should IPO now, before the market fully catches on that token costs are compounding rapidly (doubling every 45 days) while downstream productivity gains stay flat.

Chamath recounts a conversation with his CTO revealing token costs doubling every 45 days against only ~5% productivity gains, arguing this looming reckoning is why AI labs should IPO now, before investors catch on.

transcript

Chamath Palihapitiya: Right now, our token costs are doubling every 45 days. And I was like, 'Gh.' And he said, 'Yeah.' And I said, 'Well, what is the downstream productivity?' And he said, maybe 5% max. And I said, okay, so my costs are doubling every 45 days. My upside is essentially flat.

Anecdote◆

Companies like OpenAI and Anthropic should IPO now, before rising token costs and vanishing marginal returns on model improvements become visible to the market — because within a few years every AI company will face this reckoning.

Chamath recounts his CTO telling him token costs are doubling every 45 days while productivity gains are flat, and argues this coming 'reckoning' is exactly why AI labs should IPO now while the numbers still look great.

transcript

Chamath Palihapitiya: Right now, our token costs are doubling every 45 days... my costs are doubling every 45 days. My upside is essentially flat... I suspect that if you can get out now, you should get out now before all of that starts to seep into the water table because I think that's probably what allows you to get out at a huge price.

explains mechanism · 1extends · 2rebuts · 4supports · 3

Prediction◆

Because token costs are doubling every 45 days while downstream productivity gains are only about 5%, AI labs are approaching a reckoning where costs vastly outpace benefits — so a lab that can IPO now should do so before that reckoning becomes visible to the market.

Chamath recounts a conversation with his CTO showing token costs doubling every 45 days for only ~5% productivity gain, and argues every company will hit this reckoning within a few years — so labs should IPO before that becomes apparent.

transcript

Chamath Palihapitiya: I don't know how many other companies will actually go through this reckoning now, but the point is everybody in the next three or four years will for sure go through it. So, I suspect that if you can get out now, you should get out now before all of that starts to seep into the water table because I think that's probably what allows you to get out at a huge price and raise a huge amount of money.

extends · 1supports · 1

Data

SpaceX's IPO — raising $75 billion at a $1.75 trillion valuation and trading up 25% on roughly $35 billion of forward revenue — was a textbook success that both Anthropic and OpenAI studied closely, setting a blueprint that could support blockbuster IPOs at even higher valuations given their much larger revenue trajectories.

Brad Gerstner details how SpaceX's $1.75T IPO was executed flawlessly and argues Anthropic and OpenAI, watching closely, could follow with even bigger blockbuster listings given their higher revenue growth.

transcript

Brad Gerstner: It was a hugely successful IPO. They raised $75 billion at 1.75 trillion. Okay, so it went out below where we are today. It's up 25%. You know, and let's call it on 35 billion of forward revenue. So if you think about that revenue multiple, it's trading at 2 trillion on roughly 35 billion of of forward revenue. It's an incredible achievement. I think it was textbook.

extends · 1

Fact

The SpaceX IPO was a textbook success — raised $75 billion at a $1.75 trillion valuation and now trades at a $2 trillion market cap on ~$35 billion of forward revenue — and it provides the blueprint Anthropic and OpenAI will follow.

Brad Gerstner argues SpaceX's massive, well-executed IPO — including innovations on lockups and index inclusion — gives Anthropic and OpenAI a proven playbook to follow for their anticipated public offerings.

transcript

rebuts · 1

Prediction

Anthropic's IPO could be a blockbuster following the blueprint SpaceX set, given its revenue trajectory dwarfs SpaceX's at time of listing.

Brad Gerstner argues SpaceX's textbook trillion-dollar IPO (raising $75B at $1.75T) set the template on pricing, index inclusion, and lockups, and that Anthropic — rumored to be trending toward $100B+ revenue versus SpaceX's $35B — would be an even bigger blockbuster.

transcript

Brad Gerstner: You heard from Gavin, Anthropic's rumored to be, you know, trending over a hundred billion in revenue compared to the 35, right? If they exit the year at 100, that means their cap revenue next year could be well over a hundred. So, based on the SpaceX success, I think it would be a blockbuster IPO.

extends · 1supports · 2

Data

Once you strip out Nvidia's chip-sales revenue, the actual measurable EPS/ROI lift from AI across the broader S&P 493 has been roughly 0-2%, not the eye-popping numbers often cited.

Chamath describes interrogating an AI model about S&P 500 earnings growth attributable to AI, finding the headline 50% figure was inflated by Nvidia's chip sales, and that the real EPS growth for the S&P 493 (ex-Nvidia-adjacent names) was mostly pricing power and buybacks, leaving true AI ROI near zero.

transcript

Chamath Palihapitiya: I asked, what is the lift of the S&P 500 earnings per share growth since 2024 from AI? And they answered, oh, it's 50%. So then I looked through it... I asked a different question, which is, then what was the EPS growth of the S&P 493? And the answer was 9%... the actual ROI was somewhere between zero and 2%.

provides context · 1rebuts · 6

Data◆

Using Claude to strip out Nvidia's chip-sales contribution, the real EPS growth of the S&P 493 (excluding the Mag 7) since 2024 was only 9%, mostly from pricing power and buybacks — meaning AI's actual measurable ROI for the broader economy is somewhere between zero and 2%, a question sophisticated investors will eventually force enterprises to answer.

Chamath used Claude itself to analyze S&P 500 earnings and found that once Nvidia's chip revenue is excluded, real AI-driven EPS growth is minimal — implying enterprise ROI on AI spend is close to zero and will eventually be scrutinized.

transcript

Chamath Palihapitiya: So I asked a different question, which is, then what was the EPS growth of the S&P 493? And the answer was 9%. And I said, 'Okay, well that's different.' And I said, 'Unpack that.' And the overwhelming majority of that was from pricing power sitting on top of inflation. And then the other 3% was from buybacks.

explains mechanism · 1rebuts · 4supports · 1

Prediction

If Anthropic and OpenAI end the year above $100 billion in revenue, they're on a trajectory that could see that revenue grow 3 to 5x again next year — a scale of value creation never before seen.

Brad Gerstner predicts that if Anthropic/OpenAI exit the year above $100B in revenue, they could 3-5x that again next year, an unprecedented leap in Silicon Valley history.

transcript

Brad Gerstner: So let me be provocative here. If these guys end a year over a 100red billion, I think that they're on a revenue trajectory that they could 3 to 5x again next year. We've never seen anything like this. Never.

extends · 1rebuts · 2supports · 1

Claim

After Meta's 'scorch the earth' open-source strategy stumbled, Zuckerberg pivoted to a price war, publicly claiming Meta's new model matches frontier quality at roughly 1/100th the cost.

Brad Gerstner notes that after Meta's initial open-source strategy misfired, Zuckerberg shifted to waging a price war, claiming comparable quality to frontier models at a fraction of the cost.

transcript

Brad Gerstner: The meta thing was really intense because I thought okay you know we talked about the game theory which was Mark should scorch the earth with open source. I think they flubbed that play but then I think he is now said he's going to create a price war.

provides context · 1

Claim

Meta fumbled its original open-source 'scorch the earth' strategy with Llama, but Zuckerberg is now pivoting Meta into a direct price war by offering comparable model quality at a fraction of the cost of frontier labs.

Brad notes Meta's original open-source strategy misfired, but Zuckerberg's new Spark model launch signals a deliberate pivot to competing on price against Anthropic and OpenAI.

transcript

provides context · 1supports · 1

Claim

Zuckerberg is launching an AI price war, pledging Meta will offer comparable model quality to frontier labs at roughly 1/100th the cost.

Brad Gerstner points to Zuckerberg's flurry of tweets announcing Meta's new Spark model, framing it as Meta pivoting from an open-source 'scorched earth' strategy to an explicit price war promising near-frontier quality at a fraction of the cost.

transcript

Brad Gerstner: I think he is now said he's going to create a price war and so if you look at the tweet or the quote... he was basically like hey guys I'm going to give you the same quality at like 1/100th of the cost. Now again, there's a lot between here and there.

extends · 1gives example · 1supports · 1

Claim

China's strategy is to keep its AI models open-source until they catch up to the frontier, and then switch them to closed-source to capture the value for themselves — mirroring exactly what Sam Altman did at OpenAI.

Sacks argues reports of China restricting access to its top AI models reflect a broader pattern: labs like Bytedance, Alibaba (Qwen), and Zhipu (GLM) stayed open only until nearing the frontier, then closed up — the same playbook Sam Altman ran when OpenAI went from nonprofit/open to for-profit/closed.

transcript

David Sacks: I think the tactic is you stay open until you catch the frontier or you get close to it and then there's a really compelling incentive to go close because you want to capture all the value for yourself — which by the way is exactly what Sam Altman did famously at OpenAI.

explains mechanism · 1provides context · 1rebuts · 1

Claim

Chinese AI labs follow a consistent playbook: stay open source while catching up to the frontier, then go closed once they close the gap, in order to capture the value for themselves.

David Sacks explains reports of China restricting overseas access to its top models fit a broader pattern where Chinese labs (ByteDance, Alibaba's Qwen, Zhipu's GLM) go open only until nearing the frontier, then shift closed — mirroring OpenAI's own earlier pivot.

transcript

David Sacks: They I think are going closed too after having been open. And so this is I think the tactic is you stay open until you catch the frontier or you get close to it and then there's a really compelling incentive to go close because you want to capture all the value for yourself.

Mechanism

Several Chinese AI labs (Alibaba's Qwen, Zhipu's GLM) that were open source are moving to closed models now that they've caught up to the frontier, following the same playbook OpenAI itself used — stay open to build a developer community and catch up, then close off once you're competitive to capture value.

Sacks explains reports that China may restrict access to its top AI models fit a broader pattern: labs (Chinese and American alike) go open to catch the frontier, then close once they get there to capture value.

transcript

Sacks: They I think are going closed too after having been open. And so this is I think the tactic is you stay open until you catch the frontier or you get close to it and then there's a really compelling incentive to go close because you want to capture all the value for yourself.

rebuts · 2

Definition◆

Trump accounts give every American child $1,000 at birth in a privately-owned, no-cost investment account that goes into the S&P 500 and can compound for life, with contributions from family and employers.

Brad Gerstner explains the mechanics of Trump accounts: a $1,000 government seed at birth invested in the S&P 500, growing to roughly $50,000 by age 18 with modest weekly contributions, launched July 4th with over a million accounts and a billion dollars deposited in the first 24 hours.

transcript

Brad Gerstner: $1,000 for every child at birth that could compound for their life in a privately-owned investment account. So you're born, you get a social security number and you get an investment account. And if you do that, you start with $1,000 and somebody matches that and you save 10 bucks a week, that's $50,000 at age 18.

Mechanism◆

Trump accounts give every child $1,000 at birth in a privately-owned investment account that compounds tax-free in the S&P 500 for life, potentially reaching $50,000 by age 18 with just $10/week in added savings.

Brad Gerstner explains the mechanics of Trump accounts (Invest America accounts): a $1,000 government seed at birth invested in the S&P 500, free to hold for life, that could grow to $50,000 by age 18 with small regular contributions.

transcript

Brad Gerstner: So, as you guys know, the idea was very simple. $1,000 for every child at birth that could compound for their life in a privately owned investment account. So you're born, you get a social security number and you get an investment account. And if you do that, you start with $1,000 and somebody matches that and you save 10 bucks a week, that's $50,000 at age 18.

Definition◆

The Trump/Invest America accounts give every child $1,000 at birth in a privately-owned, S&P 500-invested account that compounds tax-free for life, so that adding just $10 a week can grow to roughly $50,000 by age 18.

Brad Gerstner lays out the basic mechanics of Trump accounts: a $1,000 government seed invested in the S&P 500 at birth in a free, privately-owned account that can grow to about $50,000 by age 18 with modest ongoing contributions.

transcript

Highlight slides

The Widening AI Cost Gap✦ from: AI companies should IPO now, before the market fully catches on that token costs are compounding rapidly (doubling every 45 days) while downstream productivity gains stay flat.IPO Now, Before the Reckoning Hits✦ from: Companies like OpenAI and Anthropic should IPO now, before rising token costs and vanishing marginal returns on model improvements become visible to the market — because within a few years every AI company will face this reckoning.Why AI Labs Should IPO Now✦ from: AI companies should IPO now, before the market fully catches on that token costs are compounding rapidly (doubling every 45 days) while downstream productivity gains stay flat.Every AI Company Faces This Soon✦ from: Companies like OpenAI and Anthropic should IPO now, before rising token costs and vanishing marginal returns on model improvements become visible to the market — because within a few years every AI company will face this reckoning.Chamath's Advice✦ from: Companies like OpenAI and Anthropic should IPO now, before rising token costs and vanishing marginal returns on model improvements become visible to the market — because within a few years every AI company will face this reckoning.Chamath's Call: IPO Now✦ from: AI companies should IPO now, before the market fully catches on that token costs are compounding rapidly (doubling every 45 days) while downstream productivity gains stay flat.The Widening Gap: Token Costs vs. Productivity✦ from: Because token costs are doubling every 45 days while downstream productivity gains are only about 5%, AI labs are approaching a reckoning where costs vastly outpace benefits — so a lab that can IPO now should do so before that reckoning becomes visible to the market.Why IPO Now, Before the Reckoning Hits✦ from: Because token costs are doubling every 45 days while downstream productivity gains are only about 5%, AI labs are approaching a reckoning where costs vastly outpace benefits — so a lab that can IPO now should do so before that reckoning becomes visible to the market.Claude Analysis Exposes AI's Real ROI✦ from: Using Claude to strip out Nvidia's chip-sales contribution, the real EPS growth of the S&P 493 (excluding the Mag 7) since 2024 was only 9%, mostly from pricing power and buybacks — meaning AI's actual measurable ROI for the broader economy is somewhere between zero and 2%, a question sophisticated investors will eventually force enterprises to answer.A Reckoning Is Coming✦ from: Using Claude to strip out Nvidia's chip-sales contribution, the real EPS growth of the S&P 493 (excluding the Mag 7) since 2024 was only 9%, mostly from pricing power and buybacks — meaning AI's actual measurable ROI for the broader economy is somewhere between zero and 2%, a question sophisticated investors will eventually force enterprises to answer.The Reckoning Ahead✦ from: Using Claude to strip out Nvidia's chip-sales contribution, the real EPS growth of the S&P 493 (excluding the Mag 7) since 2024 was only 9%, mostly from pricing power and buybacks — meaning AI's actual measurable ROI for the broader economy is somewhere between zero and 2%, a question sophisticated investors will eventually force enterprises to answer.Trump Accounts: $1,000 at Birth✦ from: Trump accounts give every American child $1,000 at birth in a privately-owned, no-cost investment account that goes into the S&P 500 and can compound for life, with contributions from family and employers.Trump Accounts: $1,000 at Birth, Invested for Life✦ from: Trump accounts give every child $1,000 at birth in a privately-owned investment account that compounds tax-free in the S&P 500 for life, potentially reaching $50,000 by age 18 with just $10/week in added savings.Growth to Age 18✦ from: Trump accounts give every American child $1,000 at birth in a privately-owned, no-cost investment account that goes into the S&P 500 and can compound for life, with contributions from family and employers.Growth Potential by Age 18✦ from: Trump accounts give every child $1,000 at birth in a privately-owned investment account that compounds tax-free in the S&P 500 for life, potentially reaching $50,000 by age 18 with just $10/week in added savings.Launch Momentum✦ from: Trump accounts give every American child $1,000 at birth in a privately-owned, no-cost investment account that goes into the S&P 500 and can compound for life, with contributions from family and employers.Trump Accounts: How They Work✦ from: The Trump/Invest America accounts give every child $1,000 at birth in a privately-owned, S&P 500-invested account that compounds tax-free for life, so that adding just $10 a week can grow to roughly $50,000 by age 18.Growth to Age 18✦ from: The Trump/Invest America accounts give every child $1,000 at birth in a privately-owned, S&P 500-invested account that compounds tax-free for life, so that adding just $10 a week can grow to roughly $50,000 by age 18.

Related episodes

Video · The Cognitive Revolution

Fable's Back, AI Engineer Recap, & SambaNova

→

Video · BG2 Pod

The SpaceX IPO, Fable 5, AI Capex Update & Market Check w/ Gavin Baker, Andrew Fox & Clark Tang

→

Video · a16z Podcast

Ben Horowitz on the Global Race for Tech, Power, and Influence

→