Vitalik said in a post on X that AI 2040 and its critics hold incompatible worldviews on the speed and importance of AI progress.
He said the AI 2040 position is that, unless strong measures fully stop it, some form of superintelligence will appear by 2040 across a range of scenarios. Its critics, he wrote, argue that AI 2040 underestimates humanity’s ability to coordinate and threatens freedom, but do not treat artificial superintelligence, or ASI, itself as a concentration-of-power risk.
Open to slowing or pausing AI
Vitalik said that if he were convinced current AI is just an ordinary technology, he would be closer to the critics. If he were convinced superintelligence will arrive by default in 2030, he would be closer to the AI 2040 camp. Because uncertainty remains high, he said he is still open to slowing or pausing AI.
He also said he is uncomfortable with a position held by some large AI companies and intellectuals: that open source is harmful and that the ideal outcome is for their side to hold globally controlling dominance.
Why he backs d/acc
Vitalik said an important reason he supports the d/acc platform is that several fields are worth advancing under either worldview. He listed formal verification, cryptography, secure and open hardware, pandemic resilience, defensive biotechnology, food and basic resource security, public epistemics, and non-centralized physical security.
He added that the 2040 plan has become more supportive of open source and now includes the idea of “mutually assured compute destruction.” He described that as an improvement over allowing a small number of actors to selectively strip rights from targets they identify.
On platform design and coordination
On whether AI should be slowed or paused, Vitalik said there is no way to avoid trade-offs. He suggested setting trigger conditions in advance and becoming more open to a slowdown or pause if enough conditions are met within a defined time window.
He also said that if he were Elon Musk or Zuck, he would reshape Twitter more aggressively into a platform that helps identify and facilitate large win-win agreements of this kind and brings more people into the discussion. Still, he said that idea may also be naive. For now, he does not see any non-naive plan for managing the ASI transition, and that leaves him inclined to show some tolerance toward people who are trying.