Theory of Attractor Horizons: Form-Bound Reachability of Intelligent Organizational Forms and the Limits of Human Cognition

Abstract

The development of artificial intelligence is predominantly assessed in terms of increasing cognitive performance. Yet this comparison presupposes a yardstick that captures the diversity of possible forms of intelligence only incompletely. The Theory of Attractor Horizons advances the thesis that intelligent systems, by virtue of their respective organization, can reach different regions of cognitive attractors that may partly overlap without being wholly contained within one another. Greater performance within one organizational form does not confer general access to the organizational possibilities of other forms. The theory explains this asymmetry through the dependence of reachable attractors on available transformation pathways and examines aggregation as a mechanism that can give rise to new organizational forms while also restricting existing possibilities. From this follows the far-reaching prediction that artificial systems will develop structures of thought whose internal organization the human brain cannot itself instantiate within its biological form of intelligence, even where individual results remain verifiable or technically useful. Conversely, human organizational attractors may exist that remain inaccessible to artificial systems despite their superior cognitive performance. For alignment, this entails limits to understanding-based oversight and requirements for verifiable coexistence between different forms of intelligence.