Давос: почему лидеры AI-индустрии устроили публичную потасовку

Davos: Why AI Industry Leaders Had a Public Brawl

The World Economic Forum in Davos has always been seen as a place where billionaires in expensive sweaters discuss saving the planet between lobster breakfasts and champagne dinners. But this year, something went wrong. Instead of the usual flowery phrases about the good of humanity, the leaders of the world's major AI labs staged something more like a US election primary debate, where each side tries to land the sharpest blow on opponents. If artificial intelligence used to look like a closed club of gentleman scientists, now it's a full-blown war for survival, where reputational daggers are being drawn.

The loudest episode was sparked by Demis Hassabis, head of Google DeepMind. In an interview, he didn't miss the chance to comment on rumors that OpenAI was planning to test advertisements in ChatGPT. His remark was a textbook example of passive aggression: he suggested that Sam Altman and company must really need money to make such a move so early. This is not mere gossip. In a world where training a single model costs hundreds of millions of dollars, a hint about a competitor's financial troubles is an attempt to scare off investors and poach talent. Hassabis effectively claimed that the market leader's business model is cracking at the seams.

The fun quickly spread when Dario Amodei, who runs Anthropic, joined in. His company has always positioned itself as a safer alternative to OpenAI, and Davos became the perfect platform to cement this image. When the heads of three major labs—DeepMind, Anthropic, and OpenAI—end up in one snowy town, the masks of friendliness slip off. We're witnessing the end of an era of romanticism in AI, when everyone pretended they were working for a common future. Now it's serious business, where market share matters more than ethical declarations. Lab leaders are acting like presidential primary candidates: they're not just selling their product, they're trying to destroy their neighbor's reputation.

Why is this happening right now? The answer is simple: money is running out faster than new technological breakthroughs are coming. Investors are starting to ask uncomfortable questions about monetization and return on their billions. OpenAI, being the pioneer, takes the brunt of the blows. If they introduce ads, it will destroy the user experience that everyone has gotten used to. If they don't—they might not be able to cover their bills to Microsoft for cloud computing. Hassabis and Amodei understand this perfectly and strike at the most painful spot, trying to position their companies as more stable and cleaner players compared to the "desperate" leader.

It's interesting to observe how the rhetoric in the industry is changing. Previously we heard about the risks of existential threats and superintelligence that could take over the world. Now we hear about advertising revenue, user acquisition cost, and quarterly reports. This grounding of the industry in earthly reality testifies to its maturity, but at the same time signals the loss of that very aura of mystery that helped attract capital without too many questions. When geniuses start arguing like shopkeepers, the magic disappears. But it's in such disputes that the real structure of the future market is born, where not the smartest will win, but the shrewdest and most financially stable.

Davos showed that unity in the AI community is a myth created for regulators. Behind closed doors and on panel discussions, the struggle is over who will become the face of progress in the eyes of governments. If Google manages to paint OpenAI as a greedy corporation willing to sell user attention to advertisers, and Anthropic secures its status as cautious intellectuals, the industry landscape could change by the end of the year. We're entering a phase where PR strategies are becoming just as important as neural network architecture.

The bottom line: Friendship is over, the scramble for budgets has begun. Who will be the first to admit that AI is ultimately about money, not saving the world?