Anthropic Eyes a $350 Billion Valuation: Claude Is Worth More Than You Think

It looks like the world of large language models has entered an era in which a billion dollars is no longer a serious number. Anthropic, a company founded by former OpenAI staff, is now discussing a valuation of $350 billion (3500). If you thought OpenAI's $157 billion was a lot, Dario Amodei and his team have decided to prove that investor appetite knows no limits.

This is not just an injection of fresh capital but a large-scale restructuring of ownership, in which employees will be able to cash out their stakes amid record interest in the Claude model. The context of the deal looks even more ironic given the company's origins. Anthropic was created as a "safe" alternative to OpenAI, with a focus on ethics and so-called constitutional AI.

But the reality is that training models at the level of Claude 3.5 Opus takes colossal computing power and even more colossal sums of money. To compete with Microsoft and Google, it is not enough to be "the good guys"; you need a checkbook that never runs out.

The current $20 billion (200) round is an attempt to cement Anthropic's status as the main rival to GPT-4 while other players try to catch a departing train. Why are investors willing to pay this much right now? The answer lies in performance.

Recent versions of Claude have shown that Anthropic can ship a product that works more reliably and more logically than OpenAI's offerings. Among developers, Claude 3.5 Sonnet has become the de facto standard for writing code and doing complex text analysis.

While Sam Altman busies himself with media projects and global alliances, Amodei is methodically improving the architecture, showing that a focus on quality can be worth more than aggressive marketing. The market believes Anthropic is not merely a "number two" but a potential leader of the next technology cycle. Notably, part of the deal involves a secondary sale of shares by employees.

That is an important signal for the whole industry. When a startup lets its people cash out at this valuation, it means the company is confident in its future and does not fear an exodus of staff. It is also a way to retain top talent who might otherwise leave to start their own projects.

A real war for talent is under way in Silicon Valley, and the chance to turn a paper fortune into real money is the best argument in the fight for world-class engineers. Still, it is worth asking how justified such multiples are. A $350 billion valuation puts Anthropic alongside the world's largest corporations, which have real revenue, factories and millions of paying customers.

AI startups so far have only promises to change the world and electricity bills from Nvidia. If this bubble bursts, it will drag the entire tech sector down with it. But while the money keeps flowing, nobody wants to be the last one to turn off the lights in the server room.

The belief that AGI will pay back any investment remains the main engine of the modern economy. The bottom line: Anthropic is finally turning from a research project into a financial monster. Will the company be able to keep its "safe AI" principles under pressure from investors who expect a return on $350 billion?