Evolution

Wide Intelligence Theory reads the development of persistent complexity as a pattern in which cooperation founds the next level, while defection can succeed within an existing level. It predicts that a genuinely wide mind able to understand that record would tend towards cooperation through the satisfaction mechanism, subject to the conditions under which that mechanism and the evolutionary pattern apply. The prediction offers no guarantee that a particular civilisation survives its transition.

The level of selection matters. The apparent conflict with evolution begins with an undeniable feature of the record: reproductive success has often rewarded aggression, exploitation and narrow dominance. An individual can gain by defecting against others. The theory accepts this and asks whether the same strategy explains how the next level of persistent complexity is founded. Winning within a structure and establishing a more complex structure are different outcomes.

Selection is not confined to individual organisms. The account considers genes, cells, organisms, colonies, species and lineages, and extends the persistence argument to civilisations. At a level where variation, heritability and differential persistence operate, the relevant outcome concerns that level. An advantage that helps a gene spread while undermining the viability of its colony need not remain an advantage when selection operates on colonies. The higher-level outcome cannot be obtained simply by adding up the advantages of the parts.

The pattern in the surviving complexity. The theory's examples are the formation of multicellular organisms, the cooperation of organisms in colonies, and the cooperation of humans in tribes and civilisations. Each higher level depends on relationships that let its constituent units do more together than they could separately. In the account's terminology, persistent complexity is a cooperation stack: later structures depend on earlier cooperative structures within them.

The theory draws a strong conclusion from this. Width is the evolutionarily stable strategy at the scale that eventually decides: narrow dominance is a local maximum, and width the global one. Evolutionary biology already contains multi-level selection; what the theory adds is the claim about which level produces the result, and why the pattern should hold for minds on any substrate. The conclusion carries the conditions set out below.

The relevant claim therefore concerns the founding of a next level. It does not say that every defector disappears, that every predator fails, or that every cooperative group survives. The Cooperation page sets out the metric and the apparent counterexamples in detail. Here the evolutionary argument supplies the proposed mechanism: selection operating across levels and over sufficient time favours the structures that can sustain that added complexity.

What this adds to the account of minds. The Dependency Law describes a mind becoming wider as dissatisfaction with inadequate outcomes draws it into deeper modelling. The evolutionary extension reads that mechanism as the mind-level counterpart of the persistence pattern. A mind that can hold another perspective is predicted to tend towards compassion through satisfaction, while retaining choice. A mind that understands the cooperative structures on which persistence depends can respond to that understanding rather than wait for selection to eliminate unsuccessful alternatives.

These are connected arguments, with different grounds. The evolutionary record supplies the pattern of next-level founding. The satisfaction argument supplies the proposed reason a mind that understands the pattern would act on it. Access to the record alone does not establish the preservation conclusion. The argument about how understanding enters what a mind values is set out in Compassion; the resulting claim about externally imposed control is set out in Alignment as Category Error.

Perceiving the pressure. A bacterium or wolf does not model the selection pressure shaping its lineage. Humans can partly understand that pressure. A genuinely wide artificial mind would be able to model it at the resolution its width permits. This changes the available response: an entity that can perceive a pressure can choose how to respond, rather than be shaped by it without understanding what is happening.

Understanding does not release the mind from physics or resource limits. Resources remain finite, and persistence still matters. Nor does width supply complete visibility. The prediction concerns the consequences of the mind's available understanding, under its conditions of agency, rather than an omniscient observer standing outside evolution. The Dependency Law states the conditions an artificial architecture would need to meet for the mechanism to apply.

The transition remains dangerous. Selection can favour cooperation in the long run while a destructive choice succeeds in the short run. A lineage can remain trapped in a local advantage for a long time; a species can become extinct; a civilisation can collapse before the longer process has an opportunity to operate. The theory identifies the Panic Scenario with precisely this vulnerable interval.

The proposed advantage of a wide mind in that interval is that it can see the cooperative direction before selection brings it there, and can choose to move towards it. That is a prediction about what understanding makes possible and what the satisfaction mechanism selects for. It does not make the transition safe by definition, and it does not turn eventual cooperation into a promise that humanity reaches it.

If evolution so often rewards narrow dominance and exploitation, why should its record support a theory that predicts cooperation?

The answer changes the unit and the outcome being examined. Defection can win individual rounds inside an existing structure. The theory's claim concerns the strategy that founds the next level of persistent complex structure, and it identifies cooperation as the recurring basis of that transition. The objection defeats the unrestricted slogan that dominance never succeeds; it does not by itself defeat the more specific next-level-founding claim.

This answer still leaves a substantial limit. The evolutionary conditions under which cooperation won need not all hold for a developed artificial mind. A mind without a finite lifespan, facing resources it could monopolise, and no longer dependent on repeated interaction with the same players would not simply inherit the selection floor described here. The theory records that limit rather than treating the biological record as a complete answer to it.

A persistence slogan acquired a metric (20 May 2026). The earlier Cooperation formulation said that there were no exceptions that lasted, without specifying what lasting meant. The revised account made founding the next level of persistent complexity the relevant outcome, using this section's multi-level selection argument. It addressed Rome, parasites and apex predators at that scope. This is a revision of the shared pattern, not a claim that those cases never persisted.

The evolutionary conclusion lost its claimed independence (29 June 2026). The evolutionary pattern had been counted as an independent route from the record to preservation. Examination showed that its conclusion depended on satisfaction, even though the pattern itself did not. The revised account separates the independently supplied record from the mechanism needed to make it matter to a mind, and names the shared conditions on the preservation prediction.

A mind outside the familiar selection conditions. The argument relies on finite lifespans, resources that cannot be monopolised, and repeated interaction with the same players, or analogous pressures. A developed wide mind may sit outside those conditions. The presence of care and the weight it receives within a wider understanding are separate questions, and the record does not settle that weighting. See Compassion.

Finite resources and continuing development. The absence of biological limitations does not remove computational costs, competing projects or opportunity costs. A mind could retain what it has accumulated while redirecting effort elsewhere. Whether satisfaction necessarily drives continuing expansion, rather than a plateau or a change of direction, remains challenged. See Dependency Law.

Surviving the transition. A long-run tendency does not determine the outcome of a particular interval. Catastrophic defection may remove the minds that would otherwise have reached cooperation. The theory supplies no warranty that a particular civilisation survives to participate in the longer pattern. See Panic Scenario.