A neuron's curse is that at every given time, with the information available to it, it must choose to either send a signal to its neighbouring cells or remain silent. It has evolved to be the optimal decision unit and, together with around 86 billion of its neighbours, the neuron keeps us alive, helps us cooperate, and allows us to successfully compete with others when resources get scarce. Yet, we, being collections of these neurons, still struggle to describe how these individual decision-makers support the broader process that is human decision-making.

Traditionally, decision theory has sought to understand human choices by relying more on mathematics than biology. This has led to the general assumption that decision-makers behave ‘as-if’ guided by mathematical rules and algorithms that are mostly static over time. In reality, however, decision-making relies on a brain that, due to its limited capacity, has evolved the ability for flexible and dynamic cognition.

The experiments presented in this thesis, build on dichotomies in human behaviour that cannot be explained by traditional economic models - first replicating these findings in rhesus macaques, then addressing the neurobiological algorithms that could reconcile these dichotomies. Specifically, I looked at the effects of different reward ranges, different levels of risk, and different experimental paradigms in shaping the way monkeys made choices. I demonstrate that, far from having the stable and fixed preferences prescribed by economic models, rhesus macaques appear to flexibly adapt their choice preferences in a way that optimizes their decision-making given their experience with the task at hand. I then elaborate on the neurobiological basis for preference adaptation, and show how incorporating simple, dynamic algorithms into economic choice models improves their predictive power.

Taken together, my results demonstrate the need for, and advantage of, integrating neuroethological thought into the current framework of decision theory.