gradual disempowermentIn brief · · · Written with AI
Ezra Klein warns that AI's success has a price too: human control
In a New York Times column, he argues that even if AI works as its builders hope, society could hand it control bit by bit.
In an opinion column published in The New York Times on 18 October, Ezra Klein asks what would happen if AI worked as its builders hope, rather than as they fear. In his view, that success would have a cost too: society would trade comprehension and control for prosperity, and control, once lost, doesn't come back.
Klein finds so-called gradual disempowerment more likely than a sudden loss of control. Companies, governments and labs would keep handing decisions to AI because otherwise their competitors would pull ahead. His examples are the more than 700 math manuscripts OpenAI released at once and the assistants that act on their users' behalf.
Klein says he is sceptical of a jobless future. What worries him is losing the sense of purpose and of control.
Sources
Written by AI from these sources.
- This Is the Good Scenario?Accessed 22 October 2026 · Image: The New York Times
