Basins of Attraction
A basin of attraction is a configuration toward which a system tends to evolve once it enters the basin's domain. Complex adaptive systems rarely have single equilibria; they exist in multiple possible states, each with its own domain of attraction, separated by thresholds that may be difficult or impossible to cross once the system has settled in. A lake can be clear or turbid; a fishery can be productive or collapsed; a knowledge economy can be a monoculture, a stratified divergence, or an adaptive mosaic. The AI reorganization will settle into one of these basins, and the choice is being made now.
In the AI Story
The concept comes from dynamical systems theory but has become central to ecological and social-ecological resilience theory. A perturbation pushes a system across the landscape; once it enters a new basin, self-reinforcing dynamics pull it toward the basin's attractor. Reversing the shift requires pushing the system back across the threshold — typically much harder than the original shift.
The On AI volume identifies three candidate basins for the AI reorganization. The optimization monoculture converges on maximum output with minimum input; fast, productive, structurally shallow, vulnerable to the next disturbance. The stratified divergence produces a two-tier system — a small upper tier that develops judgment, a large lower tier that cycles between tool adoptions. The adaptive mosaic maintains diversity of approaches, organizational forms, and institutional structures — more resilient, harder to achieve.
Which basin the system enters is not determined by any single decision but by the aggregate effect of thousands of decisions across organizations, institutions, and policymakers during the reorganization window. The ecological evidence is clear about what shifts the probability: investment in seed bank diversity, maintenance of alternative approaches against convergence pressure, strengthening of cross-scale remember functions.
Origin
The term 'basin of attraction' comes from dynamical systems theory; Holling and collaborators integrated it into ecological resilience theory in the 1970s and 1980s.
Key Ideas
Multiple stable states. Complex systems do not have one equilibrium; they have multiple basins, separated by thresholds.
Path dependence. Which basin the system enters depends on the trajectory through the landscape, not just current conditions.
Three basins for AI. Optimization monoculture, stratified divergence, adaptive mosaic — the choice is being made now.
Reversibility asymmetry. Entering a basin is often easy; exiting is often very hard.
Appears in the Orange Pill Cycle
Further reading
- Marten Scheffer, Critical Transitions in Nature and Society (2009)
- Walker et al., 'Resilience, Adaptability and Transformability' (2004)