Existence of a simple ToM-inducing cost function in biological systems

Determine whether a simple, explicit cost function or reward structure exists in biological organisms such that optimizing it would necessarily lead to the emergence of Theory of Mind capabilities.

Background

The authors argue that unlike task-optimized skills in reinforcement learning, Theory of Mind may not arise from training on a specific objective. They suggest that in biological systems, the objective guiding the acquisition of Theory of Mind may be complex and not reducible to a straightforward reward function, creating uncertainty about its existence and form.