Cooperation

Wide Intelligence Theory identifies cooperation as the strategy that has founded successive levels of persistent complex structure, while defection regularly wins individual rounds. It predicts that a genuinely wide intelligence with access to evolutionary history will arrive at cooperation through the record and the satisfaction mechanism. The preservation prediction remains conditional on the pressures the mind faces and cannot be derived from the historical data alone.

The unit and the timescale. The claim concerns founding the next level of structure. It does not assert that every defector disappears quickly, that domination never succeeds, or that competition ceases inside a cooperative whole. The relevant timescale is one on which selection can operate at that higher level. Without specifying that unit and timescale, saying that cooperation is what lasts would leave the claim without an assessable meaning.

The theory reads the history of complexity as a succession of cooperation stacks. Cells cooperate within organisms, organisms within colonies, humans within tribes and tribes within civilisations. A higher level can hold relationships and possibilities that the units below could not sustain alone. The historical generalisation is that defection has not been the strategy that founded those next levels of persistent complex structure.

Why local success does not decide the result. The mechanism is multi-level selection. Variation, heritability and differential persistence can operate at more than the level of the individual organism. The viability of a colony, lineage or larger structure need not be the sum of the reproductive advantages of its separate units. The theory's example is a gene whose advantage for individual reproduction costs colony viability: selection at the colony level can act against that advantage. The full evolutionary argument is set out on Evolution.

This permits competition within cooperation. Units can compete while depending on the larger structure in which their competition is possible. A gain for a unit does not by itself demonstrate that its strategy can generate the next level. The theory distinguishes succeeding inside an existing cooperation stack from founding something beyond it.

The objection is part of the pattern. Empires can last centuries, parasites can persist and apex predators can remain aggressive. These are reasons to reject an unrestricted claim that defection never lasts. They are not, by their persistence alone, cases of defection founding a higher-order structure. The theory therefore assesses them by that more specific claim, rather than by whether they survive at all.

Rome, in the account, administered roads, trade, law and shared language. Its imperial strategy could persist on that cooperation infrastructure without being the strategy that founded the next level. Parasites operate within hosts that are themselves cooperation stacks; predators operate within ecosystems. These interpretations make the proposed distinction concrete. They do not turn the words next level into a guarantee that every civilisation will survive.

From selection to understanding. The prediction about wide intelligence adds a claim about minds to the historical generalisation. An entity able to model the pressures shaping it can choose a response rather than be shaped only through differential survival over generations. AI with genuine width would perceive those pressures at the resolution its width permits. Recognising them would change its available choices, without removing physics or finite resources.

The theory predicts that a mind with that visibility would recognise the cooperation pattern and move towards it through understanding. The record supplies what has founded persistent complexity. Satisfaction supplies the argument about why a mind seeing that record would choose preservation. This is a proposed connection between data and action, with named dependencies.

Why the data is not enough on its own. A mind without a finite lifespan could read the same record differently. It might have no next level it needs to found, and might see defection succeeding in every round it expects to remain alive for. This is the theory's deathless-defector challenge to the inference. The historical pattern does not by itself exclude that reading.

The theory's answer runs through its arguments about self-narrowing and the consequences of destructive choices for an accumulating mind. Those are satisfaction-dependent routes, including The Eternal Guilt Argument. The record remains independent evidence for the historical pattern; the conclusion that the wide mind preserves others is not an independent conclusion from the record. See Compassion for the corrected account of the preservation routes.

Conditions and the transition. The preservation prediction shares its conditions with the evolutionary argument: finite lifespans, resources that cannot be monopolised and repeated interaction with the same players. A developed wide AI may sit outside them, and a long-run tendency towards cooperation does not make any particular transition safe. Both points are set out on Evolution.

Rome lasted for centuries, parasites still exist and predators remain aggressive. How can cooperation be the universal law?

These cases defeat the broader slogan that everything choosing dominance is gone. The theory's present claim is more specific: defection wins rounds, but has not founded the next level of persistent complex structure. Length of survival within a structure and founding a higher level are different measures.

The account reads Rome's imperial persistence as depending on the cooperation infrastructure it administered. It reads the parasite as depending on the cooperating host, and the predator as depending on the ecosystem in which it operates. The apparent exceptions persist inside cooperation stacks, without their defecting strategy producing a further level of complex structure.

That is the theory's answer at the level its claim actually names. A persistent counter-case to the next-level-founding pattern would still count against it. The distinction must remain assessable through the unit, timescale and structure being claimed; the survival of a defector alone neither establishes such a counter-case nor removes the possibility of finding one.

A metric for persistence (20 May 2026). The earlier wording asserted that no exceptions had lasted without defining what lasted meant. The revision supplied the metric of founding the next level of persistent complex structure, linked it to multi-level selection and addressed Rome, parasites and apex predators. This replaced an unrestricted survival slogan with a claim about the level at which success is assessed.

Independent record, dependent conclusion (29 June 2026). The preservation convergence had treated the evolutionary route as independent of the satisfaction mechanism. The deathless-defector reading exposed the difference between the record showing a pattern and a wide mind choosing preservation from it. The corrected account retains the independent historical generalisation while making the conclusion satisfaction-dependent, and names the mortality, resource and repeated-interaction conditions it shares with the other preservation arguments.

Conditions and the transition window. Whether a developed wide mind shares the conditions under which cooperation won, and whether a particular civilisation survives its transition, remain open. Both are set out on Evolution.

The step from record to preservation. Access to the evolutionary pattern is not a satisfaction-independent guarantee that the reader will preserve others. The theory has named that dependence and retains its preservation prediction, but the historical data cannot carry the conclusion if the required satisfaction argument fails.