Our research on evolving artificial life systems has shown that diverse populations are more likely to find the optimal solution than a less diverse population. What is more, they were more likely to exhibit more “contextual” behavior (i.e., conditional behavior) when they evolved in uncertain environments (where there was a one-to-many relationship between stimulus and reward) relative to unambiguous environments (where there was a one-to-one relationship between a stimulus and reward). Their neural processes were also more complex.
