English-language podcasts

Azeem Azhar's Exponential View

Anthropic's S-1: the AI lab that became a business

30 min

Listen through the publisher, then discuss the episode here or arrange an in-person Roundtable. Your city determines where to meet.

Roundtables are designed for groups of four, with a shared subject and room for everyone to speak. Come for discussion, with no dating expectation.

For the AI Investment Brief, the team at Exponential View has distilled the S-1 into 39 charts, to help you see where the revenue comes from, where the compute goes and whether the numbers add up. I think Anthropic has crossed from lab to business: its revenue has grown roughly 31x since early 2025 while its compute costs have grown about 6x, and well over half the Fortune 500 now pays Anthropic. What the S-1 leaves open is duration: about 12 weeks of contracted revenue set against more than 12 years of signed compute, a mismatch the filing itself acknowledges. In this live breakdown of the S-1, I discuss: 00:00 Is AI a real business? 02:00 The P&L: revenue outrunning compute 04:06 Why the customer-concentration worry was overblown 05:39 Half the Fortune 500 is already paying 08:14 New models drive the revenue 12:20 Scale is paying off: revenue 31x, compute 6x 13:52 Where a dollar of revenue goes 16:20 "Irresponsibly large" compute bets 17:30 12 weeks of revenue, 12 years of compute 18:54 Can Anthropic pay for its compute? 20:36 Values over profit maximization 21:39 Who controls Anthropic: Class F shares 23:00 "One of the seminal moments in human history" 24:11 Q&A: why most compute goes to training 26:37 Q&A: is the AI race just a race for capital? A note on the numbers: at 17:56 I mention $587bn of compute commitments. That is the figure on June 30, 2026, the end of Q2 and the last date the S-1 reports in full. About $400bn of it can’t be cancelled. Deals signed in Q3 have since taken the total to roughly $775bn. Anthropic's S-1 in charts: https://www.exponentialview.co/p/anthropics-s-1-in-charts AI Investment Brief: https://investmentbrief.ai About Exponential View: Welcome to Exponential View, where Azeem Azhar examines how exponential technologies such as AI are reshaping the future. Each week, Azeem shares his analysis or speaks with an expert guest to shed light on frontier shifts across technology, business and society. To keep up with the Exponential transition, subscribe to this channel or the newsletter: https://www.exponentialview.co (https://www.exponentialview.co/)/ To keep up with our AI economy research, subscribe to https://www.investmentbrief.ai (https://www.investmentbrief.ai/)/ Where to find me: Exponential View newsletter: https://www.exponentialview.co/ Website: https://www.azeemazhar.com/ LinkedIn: https://www.linkedin.com/in/azeem/ Twitter/X: https://x.com/azeem Production by EPIIPLUS1. Production and research: Baba Films, Marija Gavrilov. Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com (https://pcm.adswizz.com/) for information about our collection and use of personal data for advertising.

Episode information is provided by the publisher. Listening takes place on the publisher’s service.

Listen at the source ↗

Featured places to meet in Palo Alto, CA

Explore these places and arrange a conversation through ThinkMatch.

City guide for Palo Alto, CA

More from Azeem Azhar's Exponential View

Azeem Azhar's Exponential View

Boom, crash or takeover: Anthropic's IPO and the future of the AI economy (ft. Noah Smith)

34 min

Anthropic's S-1 looks healthy: distributed customers, improving efficiency and no sign of an imminent crash. So I went live with Noah Smith, to discuss what happens next. His worry isn't that Anthropic blows up; it's that it succeeds too well. We weighed Google's push for recursive self-improvement, cheaper Chinese models and the chance that one lab takes a share of every market, gaining power no company has had before. And if AI-driven super-science finds patterns no human can explain, what is left for us to understand? We discussed: 00:00 Is the AI boom about to crash? 03:31 Chinese AI is Android, Claude is the iPhone 05:46 How Anthropic could out-compete its customers 08:33 Could Claude outgrow the government? 10:31 Google, OpenAI and the self-improvement race 14:45 The best model takes it all 17:07 Google is disrupting itself 20:16 "One company could rule them all" 22:47 Lessons from the Reformation 26:58 Super-science: beyond human understanding 30:31 "We summon a djinn" 33:17 Cla

Discuss this episode

Community Poll

Health technologies should demonstrate better patient outcomes before widespread adoption.

Share your view, then see how readers voted. No sign-in required. This poll is shared across stories.

Do you agree or disagree?

Choose one answer

Common Quandary

A Common Quandary to discuss

Another conversation to explore, separate from this news story.

Everyone loves the group photo except you. Can you forbid them from posting it?

At a reunion, someone takes a joyful group photo that captures the night perfectly. Everyone looks wonderful except you, caught at an awful angle. The group wants to post it publicly and promises nobody else will care.

How much control should one person have over a memory shared by many?

Explore this quandary