Gates flies to Xi Jinping: wants to personally settle the AI ​​issue

Bill Gates is going to personally meet with Chinese President Xi Jinping to discuss how to slow down the most dangerous scenarios for the development of artificial intelligence. The Microsoft founder expects that Beijing will impose restrictions if the United States does so first.

From an optimist to an anxious prophet

Back in January, Gates talked about AI quite complacently, reminds xrust. But within a few months, the tone changed: the billionaire started talking about hacking, displacement of professions and the psychological pressure of technology on users. The reason was real incidents — sites built on the developments of OpenAI, Anthropic and Meta were hacked during what were called independent cybersecurity audits. According to Gates, such cases «should shake people.»

His team is now coordinating a meeting with Xi Jinping for November — during the Microsoft founder’s next trip to China. Formally, this will be their second personal meeting: three years ago, Gates became the first foreign businessman to be received by a Chinese leader since the pandemic. This time, by Gates' own admission, Xi «may be surprised» by his persistence.

What exactly does Gates propose

class=»notranslate»>__GTAG5__ He had quite a list of ideas:

• Joint control over biothreats. Countries should monitor whether AI is capable of helping in the design of molecules and the preparation of biological attacks — according to Gates, this will not in any way hinder the development of the industry itself.
• Synchronous restrictions on dangerous models. Gates is confident that if the United States were the first to commit to containing potentially dangerous AI systems, China would not refuse to do the same — “the whole world would support it.”
• Tax on the replacement of labor with tokens. Companies that displace employees with the results of AI, according to Gates, should pay a tax — the money would go to the social security system.
• Reserve of jobs for people. Up to 40% of existing professions, he believes, should be deliberately left to humans — not out of economic logic, but for the sake of “an irreplaceable manifestation of humanity,” as he described caring for his father, who had Alzheimer’s.

Separately, Gates called for the creation of a new international structure for the governance of AI — with elements of the nuclear inspection regime, international aviation rules and the Montreal Protocol on the ozone layer.

Why is this not just an American-Chinese story

Talks about the “IAEA for AI” have been going on in the industry for several years: a similar idea was promoted back in 2023 by OpenAI founder Sam Altman and the head of Google DeepMind Demis Hassabis, who compared himself to Oppenheimer. Gates himself drew a parallel between AI and nuclear technology back in 2019. The problem is that the IAEA analogy already has a weak point: the agency does a good job of inspecting nuclear plants, but has a hard time keeping violators like Iran or Russia from circumventing the rules—and the AI ​​model, unlike a nuclear reactor, is just code that doesn't need to be physically imported across borders.

So far in the United States, control over the distribution of powerful models — like the same Fable 5 from Anthropic — rests on the decisions of the federal authorities, while stricter regulation is gradually moving to the level of individual states. It is significant that even the developers themselves disagree: while Meta CEO Mark Zuckerberg publicly bets on the benefits of AI, Hassabis proposes creating a supervisory body modeled on the American financial regulator FINRA. Gates called this comparison “a stupid conversation” — according to him, arguing about the regulatory model now is as pointless as choosing which of two restaurants to discuss taxes in.

Gates the philanthropist against the backdrop of Gates the negotiator

In parallel with his political agenda, Gates continues his charitable line. His fund plans to allocate $200 billion by 2045—at current exchange rates, that’s about 16 trillion rubles—paying more and more attention to AI. Projects include working with Anthropic on AI pregnancy checks and with OpenAI to support healthcare workers in Rwanda. In the fall, Gates also plans to meet with British Prime Minister Andy Burnham and give a lecture on effective philanthropy to employees of Anthropic and OpenAI, who are preparing for an IPO.

There was also a reputational trail: in February, Gates canceled a trip to India after publishing materials on the Jeffrey Epstein case, saying that he did not realize the scale of the latter’s crimes when he crossed paths with him on charity projects.

Gates considers residents’ protests against data centers to be fighting the wrong problem. Much more serious, according to him, will be the moment when AI really hits the labor market — and then, the billionaire hints, there will be no time for rallies at the data center fence.

Sources:

• reuters.com
• lawfaremedia.org
• techcrunch.com

Xrust Gates flies to Xi Jinping: wants to personally resolve the AI issue