The Nature and Nurture of Cooperation
How does cooperation emerge from learning and from evolution?
Cooperation is one of the central puzzles of biology, artificial intelligence, and the social sciences. It can arise within lifetimes through learning, and across generations through natural selection. This project aims to explore human cooperative behavior by Artificial Intelligence at two complementary and intertwined levels:
-
Nature → Evolving cooperation over generations by natural selection
-
Nurture → Learning to cooperate within lifetime
The missing link - learned cooperation
Most research studies have been focussed on evolutionary explanations for the emergence of cooperation, and . \ either learning or evolution in isolation. Yet in natural systems, cooperation emerges from their interaction across two timescales.
Human cooperative behavior could be viewed as contemporary action based upon ancestral hardware. The origins of curent human cooperative behavior is a blend of events in history ranging from millions of years in the past until split seconds ago.
Rather than prescribing cooperative behavior through direct engineering, we investigate the fundamental and minimal constraints under which cooperative behavior emerges and prevails in a multi-agent ecosystem. We view life time experience (nurture) and evolutionary selection (nature) as dynamically intertwined processes with feedback loops rather than separable explanatory categories:
-
Evolution shapes learning capacities.
-
Learning reshapes ecological structure.
-
Ecological structure reshapes selection gradients.
Plasticity functions as the bridge between nature and nurture. It refers to an agent’s inherited capacity to modify its behavior to current conditions in response to experience. Evolution (nature) does not directly encode specific cooperative behaviors; rather, it shapes the degree of plasticity — the learning rate, memory capacity, exploratory bias, and adaptability of the agent. Learning (nurture) then operates within these evolved constraints, adjusting behavior to local ecological conditions. In dynamic environments, higher plasticity may be favored because it allows rapid adaptation; in stable environments, lower plasticity may be selected for efficiency and robustness. Thus, evolution shapes the architecture of learning, while learning reshapes the selective landscape. Plasticity therefore closes the feedback loop between generational selection and lifetime adaptation, making nature and nurture inseparable components of behavioral dynamics.
