The idea that humans could lose control over their future not through a sudden AI takeover but slowly, as AI systems take over more of the work, decisions and culture that societies depend on. The argument is that governments and companies serve people partly because they need people, and that this pressure weakens once they no longer do. It is a speculative and contested argument, made mainly by AI safety researchers.
Technical view
Set out by Jan Kulveit, David Duvenaud and colleagues in a January 2025 paper. They argue that the economy, culture and states stay aligned with human interests partly through explicit levers such as voting and consumer choice, and partly implicitly because they rely on human labour and participation. As AI replaces that participation, both forms weaken, and because the systems reinforce one another the loss could become effectively irreversible, even if each AI system behaves as intended. Sceptics, including the economist Tyler Cowen, doubt that competition and checks and balances would fail this way. The argument shifts attention from single systems to structural responses.
Related
Part of AI Terminology Index, section: Problems, Risks and Ambitions.