In The Australian, our co-founder Armina Rosenberg shared Minotaur Capital's view on Anthropic's filing for a record $US2 trillion Nasdaq IPO, a float whose prospectus warns AI could pose "catastrophic or existential risks to humanity". In our view, Anthropic has solved how to get customers to pay for its models, but being able to fund its own computing needs is "the real test".
"Anthropic is now one of the anchor tenants of the global AI build-out. It has signed up for more than $US800bn of compute and equipment, most of which hasn't started yet," Armina Rosenberg said.
The prospectus shows first-half revenue of $US16.3bn alongside a loss of $US80.7bn. Computing spend rose from $944m in the March 2025 quarter to $5.6bn in the June quarter. Over the same period, computing costs as a percentage of revenue fell from 251 per cent to 48 per cent. Looking ahead, the key competitive risk we see is customers deciding currently available models are good enough, making continued development academic.
Beyond the headline warnings about humanity, the filing offers a rare look at the economics of a frontier AI company. As a global equities fund manager, we focus on understanding those economics to separate AI hype from durable value.
To read the article, click the link below.