The Silent Consensus

OpenAI
The author argues against the silent consensus that humans will inevitably forfeit primacy to superintelligent machines and calls for retaining human control.

Summary

In this essay adapted from a Harvard lecture, the author critiques what he calls the silent consensus: the unspoken belief that humans will inevitably cede economic and political primacy to superintelligent AI. He describes a future where competitive pressures lead humans to delegate more and more decisions to machines, eventually becoming passengers in a world run by AI. While acknowledging that this outcome might be beneficial if machines are benevolent, he rejects the inevitability and the loss of human agency it entails. He argues that human primacy rests not on intellectual superiority but on our historical and moral inheritance—our civilization, our values, and our participation in what Michael Oakeshott called the conversation of mankind. He calls for active resistance to this consensus, including building AI systems that are accountable, transparent, and compatible with human control. He emphasizes the need for technical and legal infrastructure to ensure humans remain accountable for AI actions, and he urges a broader ambition to use AI to expand human capabilities rather than diminish them.

(Source:OpenAI)