Elon Musk, founder of xAI, has made another bold statement that raises questions about humanity's future in the era of artificial intelligence. According to his estimates, within just five years, AI could surpass the combined intelligence of all people on the planet, and within the next decade, humanity will most likely lose control over this technology. This is not just a hypothesis—it is an analysis based on the current pace of neural network development and computing power.

Musk emphasizes that AI will be capable of performing virtually any intellectual task better than humans. The only exception, he says, is "being human." He calls this scenario the "age of abundance," predicting that by 2036, people will be able to obtain virtually any goods and services thanks to the synergy of AI and robotics. However, the risks remain, and they are significant.

Safety and Mutual Checks

Musk acknowledges that stopping the development of AI and robotics is impossible—even if a "stop button" existed, pressing it would not be justified. Instead, he proposes that leading AI companies implement a system of mutual checks: before a release, provide competitors with limited access to new models to identify dangerous capabilities. If the risks are not addressed, Musk believes government regulators should step in. This idea is already gaining traction—a similar initiative was previously proposed by Google DeepMind CEO Demis Hassabis.

Criticism of OpenAI and a Look at Competitors

In the same interview, Musk criticized OpenAI's transformation from a non-profit organization into a commercial entity. He recalled that the project was created as an open company "for the world," but ultimately turned into a closed business. Regarding Anthropic CEO Dario Amodei, Musk was more restrained, calling him a "principled person," but emphasized the need for cooperation on safety issues even among competitors.

It is worth recalling that Eliezer Yudkowsky and Vitalik Buterin have also warned about the catastrophic risks of superintelligent AI. However, as Near Protocol co-founder Illia Polosukhin rightly noted, AI lacks economic incentives to destroy humanity. From my perspective, the key question is not whether AI can seize control, but whether we will have time to create adequate governance mechanisms before the technology becomes uncontrollable.