Anthropic and $300 billion: when ambitions outpace common sense

The venture capital world in AI has completely lost touch with earthly gravity. While we discuss new model features, behind the scenes a financial drama is playing out on a scale that's hard to comprehend. Anthropic, the company founded by OpenAI defectors in pursuit of "safe and ethical AI," now claims a $300 billion valuation. To put this in perspective: that's more than the market capitalization of most S&P 500 companies. Dario Amodei and his team want to raise another $20 billion, and this is happening as their previous funding rounds haven't even had time to "cool" in bank reports.

Let's recall where this all started. Anthropic always positioned itself as the "adults in the room." They left Sam Altman because they believed his approach was too commercialized and risky. They promised us constitutional AI that wouldn't be rude and wouldn't take over the world. But the irony is that building "safe" AI requires exactly as much computational power as building "dangerous" AI. And for computational power, you have to pay Jensen Huang at NVIDIA—pay in advance and pay a lot. This is where the aggressive fundraising drive comes from.

Why does a startup with no billion-dollar profits get a $300 billion valuation? It's a game of survival. In the LLM industry right now, you either build your own supercomputer the size of a small city or you're out of the game. Google and Amazon have already invested billions in Anthropic, but it's clearly not enough. The new $20 billion isn't just money for engineer salaries—it's fuel for the next leap forward. Training Claude 4 or Claude 5 will require infrastructure that previously only nations could afford. Anthropic is trying to establish itself as the third pole of power alongside Microsoft/OpenAI and Google.

Here's the interesting part: who will give this money? Traditional venture funds are already eyeing these numbers with caution. We're likely to see strategic investors or Middle Eastern sovereign wealth funds back in action. For them, $20 billion is a bet on who will own the operating system of the future. If Anthropic closes this round, it will create enormous pressure on competitors. Even Mistral or Cohere will look like dwarfs that either need to find equally generous backers or retreat into narrow niches.

But there's a dark side to this story. A $300 billion valuation puts obligations on the company that are almost impossible to meet in the short term. Claude is an excellent model—many developers prefer it for its conciseness and logic. But does it generate enough money to justify such a capitalization? Not yet. We're witnessing a classic example of bubble inflation, where an asset's value is determined not by its current revenue but by investors' fear of missing the next technological epoch. If tomorrow it turns out that model scaling has hit a ceiling, these $300 billion will turn into a pumpkin faster than you can finish typing a prompt.

The bottom line: Anthropic is finally transitioning from the league of research labs to the league of financial titans. Do they have the talent to turn these billions into intelligence that will truly change the economy, or are we just watching the most expensive arms race in silicon history?