The Geometry of $65 Billion: Why Anthropic’s Run Rate Is a Warning, Not a Victory

Silence is the loudest warning. Last week, a number surfaced through the cracks of a private investor update: Anthropic’s annualized revenue run rate had reached $65 billion by the end of July, leaving OpenAI’s $40 billion trailing in the dust. The 622% expansion across seven months feels like a heartbeat—fast, rhythmic, almost intoxicating. But as someone who spent years auditing the mathematical elegance of early Ethereum smart contracts during the 2017 ICO frenzy, I’ve learned that numbers like these often hide a deeper geometry. They tell a story not of progress, but of concentration. And in a bull market where euphoria masks technical flaws, we must see through the marketing with code-audit eyes. Context: The AI race is now a public spectacle. Anthropic filed a confidential prospectus with the SEC in June, and Bloomberg reports its Wall Street debut could come as soon as this fall, with investors expecting a $2 trillion valuation. The company’s preliminary second-quarter revenue topped $11.5 billion, against $787 million a year earlier—a 14x jump. Meanwhile, OpenAI’s run rate has roughly doubled since end of 2025 to $40 billion. Both firms are preparing for IPOs, and the narrative is clear: AI is the new gold rush. But for those of us who have watched DeFi Summer bloom and wither, the pattern is hauntingly familiar. When liquidity pools become too deep in one place, they become a single point of failure. Anthropic’s run rate isn’t just a revenue metric; it’s a measure of how much power is being funneled into a single organization. Core: Let’s dig into the numbers with the same rigor I applied to Golem’s Sybil resistance mechanisms in 2017. An annualized run rate of $65 billion means Anthropic is currently generating revenue at a pace that, if sustained, would yield that amount in a full year. The path from $9 billion at end of 2025 to $47 billion in May to $65 billion in July represents a 38% growth in just two months. Quarterly revenue more than doubled from $4.73 billion in Q1 to over $11.5 billion in Q2. These are staggering figures, but they are also fragile. Based on my experience during the 2022 bear market, when I audited governance tokens of major DAOs and found 12 critical centralization flaws, I know that exponential growth often hides brittle architectures. Anthropic’s revenue is heavily dependent on a few large enterprise contracts and the continued demand for its Claude model. If a competitor releases a superior model, or if regulatory pressure squeezes the market, that run rate could collapse faster than a poorly designed liquidity pool. The geometry of trust here is not decentralized; it’s a single point of gravity. DeFi breathes; don’t let it suffocate under the weight of centralized AI. But there’s a deeper layer. The 622% expansion in seven months is reminiscent of the DeFi Summer of 2020, when Uniswap and Compound saw liquidity multiply overnight. I co-authored a whitepaper on “Liquidity as a Public Good” during that time, arguing that composability creates organic resilience. Anthropic’s growth, however, is not organic—it’s fueled by venture capital and a narrative of inevitability. The company posted positive adjusted operating income, but that’s a vanity metric in a bull market. The real question is not how fast revenue grows, but how sustainable the underlying system is. In my work with the Beijing-based fintech lab in 2024, I used game theory to show that centralized networks can withstand institutional pressure only by compromising their core values. Anthropic’s core value is its AI model, but the company controls access, pricing, and even the ability to shut down the service. That’s not scaling; it’s slicing already-scarce trust into fragments. Contrarian: The conventional wisdom is that Anthropic’s run rate proves the superiority of its technology over OpenAI. But I’d argue the opposite: the gap between $65 billion and $40 billion is a manufactured narrative, much like the liquidity fragmentation narrative that VCs use to push new DeFi products. Both companies calculate run rate differently, and neither number is audited. More importantly, the real battle isn’t between Anthropic and OpenAI—it’s between centralized AI and decentralized alternatives. The euphoria around these IPOs blinds us to the fact that AI is becoming a monopolistic utility, controlled by a handful of players. In 2022, I worked with three DAOs to implement regenerative governance, teaching them to prune dead branches to save the tree. The same principle applies here: the AI industry needs to prune its centralization before it collapses under its own weight. The $2 trillion valuation for Anthropic is not a sign of health; it’s a sign of market fever. Silence is the loudest warning—and the silence from the crypto community about this concentration is deafening. Takeaway: Geometry remembers what markets forget. The run rate numbers will fade, but the structure of power they represent will remain. As we prepare for a decentralized future, we must ask ourselves: are we building systems that scale without centralizing, or are we simply trading one monopoly for another? The answer lies not in the run rate, but in the code. Prune the dead branches, save the tree. Let the market’s euphoria pass, and let the geometry of trust guide our hands.