​Intelligence evolved to respond to the many problems that organisms face. Accordingly, intelligence can be defined based upon the problem-solving ability of organisms in variable environments.

But the evolution of responses to variable environments does not require the cognition and learning of animals. Since the dawn of life, bacteria have dealt with fluctuating environments and life-or-death challenges. The result is a baroque network of molecular responses from the environments bacteria experience to the physiological responses that help them persist.

While scientists have mapped out the details of the many bacterial regulatory networks, the evolutionary logic underlying the resulting responses receives little attention. And yet, these responses speak to the most primitive—and alternative—forms of “intelligence” generated by natural selection.

The goal of this project is to build an evolutionary framework to understand such bacterial decision making. In particular, it will focus on one of the most important discriminations that organisms face: discerning friend from foe.

Kevin Foster’s team will combine differential equation models, game theory, agent-based models, and machine learning to study the evolution of bacterial decisions across a diverse range of simulated challenges. Importantly, they will constrain models based on the known molecular biology of bacteria, exploring what can be achieved with their regulatory networks.

The team’s goal is to publish the work in a series of high-profile papers that explore the limits—and heights—of problem solving in some of the earliest and simplest organisms on earth. They will compare the solutions achieved by bacterial regulatory systems to those achieved by models that allow the flexibility of animal intelligence. In this way, they seek to shed new light on the conditions that favor the evolution of intelligence.