Can Anthropic make enough money to justify its value? | The Indicator

Anthropic’s public financial filing is presented as a heat check on the AI economy and on whether the company’s growth can justify its valuation. The host and a guest trace how Anthropic began in 2021 as a safety-focused breakaway from OpenAI, with co-founders who were not even sure they wanted to build a company at all. That origin story contrasts with the current reality: Claude has become one of the most popular large language models, especially for coding, and revenue has surged from $41 million in 2023 to over $33 billion over the last 9 months.

The financial statements make Anthropic look unusually lean for an AI giant. It spent a little less than a billion dollars on property and equipment so far this year, far less than Meta’s $83 billion, and its balance sheet shows cash with liabilities that can be paid in stock rather than cash. But the guest argues that the filings do not tell the whole story because Anthropic has also committed to spend about $900 billion on computing, with about half of that expected in the next four years.

That leaves investors with a central question: can sales keep rising fast enough to fund the servers, chips, and cooling systems behind the model? The episode also flags the circular, intertwined relationships in AI infrastructure, including Anthropic’s $110 billion obligation to Amazon over 10 years, and asks whether the bigger prize is artificial general intelligence that can improve itself and reshape the economy.