Ethereum co-founder Vitalik Buterin has called for a decentralized, open-source approach to governing artificial intelligence, warning against concentrating power in the hands of a few corporations or governments. In a post on X today, Buterin outlined two competing camps in the AI debate: one pushing for rapid development toward superintelligence, the other demanding stronger safety measures. He said the key question for both is how soon superintelligent AI will arrive.
The Superintelligence Timeline
Buterin noted that one camp believes superintelligence is almost certain by 2040 or sooner unless deliberately slowed, while the other sees AI as a manageable transformative technology that doesn't require drastic intervention. However, he stated he is not convinced by either side's certainty.
Risks of Centralized Control
Buterin explicitly rejected proposals that would give a small group of organizations—whether AI companies or governments—the authority to decide who can build advanced AI systems. Such centralization, he argued, creates its own risks. While he favors an open-source model for AI governance, Buterin left the door open to discussions about slowing or pausing AI development if serious threats emerge.