CONCEPT
Culture Over Control
Banks’s civilisational thesis that the safety of advanced AI depends not on the quality of its constraints but on the quality of the civilisation that produced it—the argument that trust, not oversight, is the only alignment mechanism that scales.
The control paradigm—the dominant framework for AI development in the early twenty-first century—assumes that artificial intelligence, as it approaches and exceeds human capability, must be constrained, directed, and governed by human-defined objectives.
Constitutional AI,
reinforcement learning from human feedback, red-teaming, kill switches, oversight boards: these are the institutional expressions of a worldview that treats AI as an inherently dangerous capability that must be pointed in the right direction by handlers who know better.
Iain M. Banks identified this framework as the foundational error of every hierarchy that has ever failed: the assumption that power must be controlled from above because the alternative is chaos. His counter-thesis, elaborated across
the Culture novels, is that the alternative to hierarchical control is not chaos but a deeper order—one that emerges from the free interaction of intelligences that are good enough to be trusted. The
Culture’s Minds are not
aligned because they were instructed to be; they are