Why the All-In Podcast Is Warming to Anthropic: The Business Behind Silicon Valley's Changing Attitude
For much of the artificial intelligence boom, Anthropic has occupied an unusual position in Silicon Valley. It has built some of the industry's most powerful AI models while publicly warning about the risks of developing increasingly capable systems. That combination has made the company both an important competitor to OpenAI and a source of tension among technology investors who believe excessive regulation could slow the industry down.
Now, some of its most vocal critics are giving the company a different kind of attention.
In an October 30, 2026, episode of the All-In Podcast, hosts Chamath Palihapitiya, Jason Calacanis, David Sacks and David Friedberg devoted a substantial discussion to Anthropic's newly filed IPO registration statement. The conversation was striking not simply because the hosts praised the company's prospects, but because they examined Anthropic as a potentially extraordinary commercial business, even while continuing to question its approach to AI safety and regulation.
David Sacks described the opportunity as potentially one of the greatest venture outcomes in history. Chamath focused on the company's infrastructure requirements and potential valuation, while Calacanis examined its margins and the economics of the wider AI industry.
The change in emphasis reveals something important about Silicon Valley. A company's commercial success can become difficult to dismiss even when its founders, political positions or regulatory arguments remain controversial.
Anthropic is increasingly being judged on a question that extends beyond its philosophy: can it turn the extraordinary demand for artificial intelligence into a durable, profitable business?
The IPO Filing That Changed the Conversation
The immediate trigger was Anthropic's S-1 registration statement, filed with the U.S. Securities and Exchange Commission on October 26 ahead of a proposed public offering.
An S-1 is a document companies submit when preparing to sell shares to the public. It gives prospective investors a detailed account of the business, financial performance, ownership structure, risk factors and obligations.
For Anthropic, the filing offered investors a more detailed look at a company that had previously been valued largely through private funding rounds and reports about its rapidly expanding AI business.
The October 30 All-In episode was titled "Anthropic S-1 Shocker, Greatest Business Ever? Google's $500B Wipeout, Midterm Predictions." Its first major discussion focused on the filing, with the hosts examining revenue growth, reported losses, gross margins, financing and the enormous infrastructure requirements behind Claude.
The significance was not that every host suddenly agreed with Anthropic. Rather, the filing gave them a new set of financial facts to debate.
For investors, the question had shifted from whether Anthropic's approach to AI development was politically or philosophically attractive to whether its economics could support a company of extraordinary scale.
Anthropic's Growth Is Becoming Difficult for Investors to Ignore
Anthropic's commercial expansion is central to the changing conversation.
Its Claude models have become important tools for businesses, particularly in software development, coding assistance, enterprise knowledge work and increasingly complex automated tasks.
The company's IPO filing reported substantial enterprise adoption. As of September 30, 2026, Anthropic said it had more than 11,900 customers generating at least $100,000 in trailing twelve-month revenue. More than 1,900 customers generated at least $1 million, while approximately 250 generated at least $10 million.
These figures matter because enterprise customers are different from individual consumers experimenting with a chatbot.
Businesses integrate AI into existing operations, software systems and workflows. Once a model becomes important to a company's engineering, customer service or internal processes, the relationship can extend beyond a single subscription. It can become part of the organization's technology infrastructure.
This creates opportunities for recurring revenue, larger contracts and deeper customer relationships.
Anthropic's enterprise figures also indicate that its business is not dependent solely on attracting millions of casual chatbot users. It is selling access to AI capabilities that companies increasingly consider important to their operations.
For investors, that distinction is significant. Enterprise adoption can create a more substantial commercial foundation, although customer concentration, contract renewals, pricing pressure and infrastructure costs remain important risks.
The 85% Gross Margin Argument
One of the most striking parts of the All-In discussion concerned Anthropic's gross margins.
The hosts discussed a figure of approximately 85%, using it to argue that the company could have attractive underlying economics despite the enormous cost of developing and operating frontier AI models.
Gross margin measures the revenue remaining after direct costs associated with delivering a product or service have been deducted. It does not represent net profit.
This distinction is especially important in AI.
Running large models requires expensive computing infrastructure, including advanced processors, data centres, electricity and networking. Developing new models also involves research, engineering and substantial investment.
A company can report attractive gross margins while still spending enormous sums on research, infrastructure expansion, stock-based compensation and other operating expenses.
It is also important to distinguish the hosts' interpretation from an independently established company-wide profitability figure. The approximately 85% figure discussed in the podcast was presented as an incremental gross-margin estimate, not proof that Anthropic earns an 85% net profit margin.
Nevertheless, the margin argument matters because it challenges one of the biggest questions facing the AI industry: whether the cost of serving increasingly powerful models will consume most of the revenue they generate.
Anthropic's argument is that improvements in model efficiency can reduce the cost of performing individual tasks while allowing the company to retain a substantial portion of the savings.
David Sacks emphasized this relationship during the podcast. He argued that token costs were falling as models became more efficient, while margins were improving rather than collapsing.
If that pattern continues, Anthropic could benefit from both increasing demand and improving economics.
However, gross margin alone cannot establish that the company has achieved sustainable profitability. Investors must also examine capital expenditure, infrastructure commitments, cash flow and the cost of maintaining its technological position.
The $80 Billion Loss: Why the Number Needs Context
One of the most controversial aspects of Anthropic's IPO filing is its reported net loss of approximately $80.7 billion for the first six months of 2026.
At first glance, such a figure appears difficult to reconcile with the company's extraordinary revenue growth. But the accounting explanation is important.
Anthropic's filing attributes approximately $72.4 billion of the six-month loss to changes in the fair value of convertible promissory notes, derivative liabilities and other financing instruments.
These changes were primarily driven by increases in the estimated value of Anthropic as its valuation rose.
In simple terms, the accounting treatment can create a substantial reported expense when the company's valuation increases, even though that revaluation does not represent an equivalent cash payment made during the period.
This is why David Sacks argued on the podcast that the headline loss gave an incomplete picture of the company's underlying business economics.
There is a legitimate point in that argument. Investors should distinguish between non-cash valuation adjustments and the money a company actually spends operating its business.
But the opposite mistake would be to dismiss the entire loss as meaningless.
Anthropic still has substantial operating expenses, infrastructure commitments and negative cash-flow requirements. Its filing also warns that future investments could require significant additional capital.
The more useful question is therefore not whether the $80.7 billion figure is entirely representative of operating performance. It is how much of Anthropic's spending is necessary to maintain its competitive position, how quickly revenue can grow, and whether future cash generation can support its commitments.
The accounting adjustment explains much of the headline figure. It does not eliminate the underlying financial risks.
Why Compute, Data Centres and Electricity Matter More Than Ever
Anthropic's growth depends on access to computing power.
Every time a business uses Claude to analyse documents, generate code, summarize information or run an AI-powered workflow, computing resources are required. Training more capable models also demands substantial infrastructure.
The company disclosed that compute and infrastructure expenses reached approximately $8.9 billion during the first six months of 2026, compared with approximately $2.3 billion in the same period of 2025.
That represents an increase of roughly 283%.
Yet compute expenses fell from approximately 201% of revenue to 55% over the same comparison period. This reflects the fact that revenue expanded much faster than the company's previously committed computing costs.
The improvement is encouraging, but it does not mean infrastructure is becoming an insignificant expense. Anthropic expects compute and infrastructure to remain its largest operating expense for the foreseeable future.
This is where Chamath's contribution to the All-In discussion becomes particularly important.
He focused on the risks hidden behind the company's growth figures, especially the difficulty of securing enough data-centre capacity, advanced chips and electricity to support future demand.
The challenge is not simply buying more processors. AI infrastructure requires land, power connections, cooling systems, networking equipment, financing and long construction timelines.
A company can have strong demand and substantial financial resources but still struggle to deploy enough infrastructure quickly.
Anthropic's filing acknowledges these constraints and warns that its growing infrastructure investments could materially affect its cash flow.
This also explains why the AI competition is increasingly becoming an infrastructure competition. The companies that can secure reliable computing capacity may have an advantage in developing and delivering the next generation of AI models.
The $808 Billion Infrastructure Question
The scale of Anthropic's infrastructure commitments has attracted particular attention.
A WIRED examination of the October 26 filing identified approximately $417 billion in future minimum payments associated with compute and infrastructure leases, followed by another $245 billion in additional infrastructure commitments.
The filing also referred to between $125 billion and $150 billion in expanded commercial agreements that had not yet been finalized, alongside approximately $21 billion in short-term commitments.
Taken together, these figures point to potential infrastructure-related commitments of at least $808 billion if the arrangements proceed as described.
That number requires careful interpretation. It does not mean Anthropic has already spent $808 billion, borrowed that amount or owes every dollar immediately. Some of the commercial agreements were not finalized, and the commitments extend across different periods and contractual arrangements.
Nevertheless, the figures reveal the scale of the infrastructure strategy required to compete at the frontier of AI.
Anthropic needs computing capacity not only to serve existing customers but also to train future models and support the additional demand it expects from businesses.
The opportunity is enormous, but so is the execution risk. If demand grows more slowly than expected, or infrastructure becomes more expensive than anticipated, the company could find itself carrying commitments that are difficult to justify.
This is one of the most important questions investors should examine before evaluating Anthropic's proposed public offering.
Why David Sacks and Anthropic Have Been at Odds
The changing tone of the All-In discussion becomes more interesting when viewed against the history of David Sacks' criticism of Anthropic.
Sacks has repeatedly questioned the company's approach to AI regulation.
His central argument is that Anthropic's calls for stronger AI safety rules could create regulatory barriers that established companies are better equipped to handle than smaller competitors.
This is known as regulatory capture: a situation in which regulation intended to protect the public can instead become shaped in ways that benefit established businesses and make competition more difficult.
Sacks has argued that requiring extensive government approval for advanced AI models could slow new entrants while protecting companies that already have the resources, technical expertise and legal teams to comply.
Anthropic has rejected the suggestion that its policy agenda is designed to protect its market position. The company has argued that powerful AI systems create risks that justify stronger oversight and that safety requirements should apply to developers according to the capabilities and risks of their systems.
This is a genuine disagreement about how the AI industry should be governed.
It is also a disagreement with major commercial implications. If governments introduce expensive compliance requirements, the cost of developing frontier AI could rise. If regulation is too weak, however, the public may be left with insufficient protection against serious risks.
The debate cannot be reduced to whether one side supports safety and the other opposes it. Both positions involve questions about innovation, competition, accountability and who should determine the rules governing increasingly powerful technology.
The October 30 Episode Shows That the Disagreement Has Not Disappeared
It would be misleading to interpret the latest All-In episode as a complete reversal of the hosts' previous positions.
After discussing Anthropic's financial prospects, the podcast moved into a separate segment titled "AI's Biggest Problem and Anthropic's Doomer Showdown."
The discussion returned to concerns about AI safety, the political influence of technology companies and the financial consequences of the industry's enormous infrastructure spending.
Chamath raised a broader concern: the AI investment boom has become so large that major technology companies are competing with other borrowers, including the US government, for capital.
David Sacks, meanwhile, continued to question pessimistic narratives about AI and the economic assumptions behind some of the industry's safety arguments.
This distinction matters because the hosts were evaluating two different things.
The first was Anthropic as a business: its revenue, customer growth, margins and potential value as a public company.
The second was Anthropic as a political and institutional actor: its views on AI safety, government oversight and the future structure of the industry.
A company can perform exceptionally well commercially while its critics continue to disagree with its policy positions.
In fact, Anthropic's success may make those disagreements even more consequential. A company approaching the scale of a major public technology business has a much greater influence on the future of AI than a small research laboratory.
The Bigger Business Story: AI Is Becoming an Enterprise Market
The enthusiasm surrounding Anthropic also reflects a broader shift in the artificial intelligence market.
The first phase of the generative AI boom was driven largely by public attention to chatbots and the novelty of interacting with powerful language models.
Businesses are now looking beyond experimentation.
Companies are integrating AI into software development, customer support, research, internal operations and business decision-making. The next stage involves AI agents capable of completing multi-step tasks rather than simply responding to individual prompts.
This transition could significantly increase demand for computing resources.
A person asking a chatbot a question may consume relatively little computing capacity compared with an AI agent that spends several minutes analysing a codebase, testing changes, reading documents and completing a workflow.
For Anthropic, this creates an opportunity to increase revenue as customers use more advanced models for more valuable work.
It also creates competition.
OpenAI, Google, Microsoft and other AI developers are pursuing similar enterprise markets. Open-source models could put pressure on prices, while improvements in hardware and software efficiency could change the economics of delivering AI services.
Anthropic's challenge is to maintain a technological advantage long enough to establish durable customer relationships and generate returns on its infrastructure investments.
Its growth demonstrates that businesses are willing to pay for powerful AI tools. It does not guarantee that the company will retain its current position.
The IPO Raises a Larger Question About the Future of AI Investment
Anthropic's proposed public offering is important because it could provide investors with a more direct way to participate in the economics of frontier artificial intelligence.
The All-In hosts discussed a potential valuation in the range of $1.5 trillion to $2 trillion and an offering that could raise as much as $100 billion. These were figures discussed in the podcast, not final offering terms established by the preliminary SEC filing.
At such a valuation, investors would be making a substantial bet on future growth.
They would effectively be betting that enterprise AI adoption will continue expanding, that model developers can maintain pricing power, and that improvements in efficiency will allow revenue to grow faster than the cost of delivering intelligence.
They would also be accepting considerable uncertainty around infrastructure financing, competition, regulation and the pace of technological change.
The scale of the proposed offering illustrates how much financial value investors now attach to advanced AI.
It also raises questions about whether the industry's expectations have moved ahead of what its infrastructure and economics can realistically support.
Anthropic's filing provides evidence of extraordinary commercial demand, but it also reveals a business that must continue making enormous investments before all the expected computing capacity becomes operational.
For investors, those two realities must be considered together.
Silicon Valley's Changing Attitude Is About More Than Anthropic
The All-In Podcast's changing emphasis is part of a wider Silicon Valley debate about the relationship between commercial success and ideological disagreement.
For years, some technology investors viewed Anthropic's approach to AI safety as a potential obstacle to innovation. Its public warnings and support for regulation created friction with people who believed the industry should be allowed to develop with fewer restrictions.
But Anthropic's financial performance has changed the terms of that debate.
The company is no longer being discussed only as a research organization with strong opinions about the future of artificial intelligence. It is becoming a major commercial institution with a rapidly expanding customer base, enormous financial resources and an increasingly important position in the global technology industry.
That makes it harder for investors to ignore.
The latest All-In discussion does not demonstrate that the hosts have abandoned their previous criticisms. It shows that they can recognize Anthropic's business potential while continuing to challenge its views on AI governance.
Ultimately, that is the more revealing story.
Silicon Valley does not have to agree with a company's philosophy to recognize the value of its technology. And in an industry where revenue growth, computing capacity and technological leadership increasingly determine market influence, commercial performance can change the tone of even the most persistent critics.
Anthropic's next challenge will be proving that its extraordinary growth can translate into sustainable financial returns without losing the technological position that made the business valuable in the first place.
The All-In hosts may still disagree with Anthropic about how AI should be governed. But they are increasingly taking seriously what its business could become.