FT: Anthropic Needs $600 Billion in Annual Revenue to Justify Compute Spending
According to FT, Anthropic would need roughly $600 billion in yearly revenue to make its compute commitments economically viable.
FT writes that Anthropic needs to earn $600 billion a year to justify its compute spending.
Anthropic has signed compute capacity contracts worth about $1 trillion (according to an updated S-1, obligations are nearly $900 billion, but additional payments will raise the total). The obligations run until 2045, giving Anthropic time to pay them off. But how much must the business generate for such investments to be worthwhile?
Let’s calculate.
Assume a desired annual return of 20% (a reasonable benchmark for a mature business). That implies $200 billion in net profit per year. Accounting for taxes, that is $241 billion. Applying a ~40% margin, revenue would need to be about $602 billion annually.
For comparison, at the current growth pace, Anthropic’s revenue is roughly $68 billion per year, with margins around 15%. Alphabet earns $471 billion, while Apple earns $478 billion. This means Anthropic would have to become more profitable than the world’s largest companies.
That was not simply bold...
P.S. Meanwhile, Dario Amodei cheerfully opens an exchange.
Why it matters
AnalysisThe estimate shows how large AI-economy expectations are: to cover infrastructure commitments, a model company would have to become more profitable than giants such as Alphabet and Apple.
Discuss in community
Share your questions and insights with developers
Related in Ecosystem
Practical guides, core terminology, and agent skills
Scaling Laws in AI
An empirical law established by OpenAI and Google (formulated by Jared Kaplan in 2020) asserting that the performance of a language model predictably increases as a power law with the growth of three factors: the number of model parameters, the volume of training data, and the computational power expended (Compute).
Computer Use via AI
Anthropic's technology (introduced in Claude 3.5 Sonnet) enables neural networks to control operating systems like a human: taking screenshots, analyzing window coordinates, moving the mouse cursor, clicking buttons, and entering text without using specialized APIs.
Anthropic mathematician claims possible proof of Riemann hypothesis with Claude
Levent Alpoge says collaborative work with Claude may have led to a proof of the Riemann hypothesis.
Anthropic goes public with shares trading near $100
Anthropic completed an $81 billion IPO and raised another $26 billion through a private placement to large investors, with demand exceeding supply sevenfold.