Ethereum co-founder Vitalik Buterin has proposed a compromise in the ongoing debate over AI superintelligence. In a recent post, Buterin highlighted the divide between two camps: one that believes superintelligence is inevitable by 2040 and must be controlled, and another that views AI as a regular technology, opposing pauses as human control. Buterin, open to slowing AI development, criticized large AI companies for advocating against open-source, seeing it as a power grab. Buterin's compromise suggests pre-agreeing to pause AI development if extreme scenarios, such as pandemics or mass unemployment, occur. If these do not materialize, the other camp's approach would proceed. He acknowledged the imperfections in all strategies, emphasizing the importance of advancing defensive technologies like formal verification, cryptography, and secure hardware in any scenario.