Abstract
How should humanity steer the future? One interpretation of this question leads us to moral philosophy: how ought humanity act? Meta-ethics may cloud the issue, however, formalisations developed in the computer sciences that provide a framework for optimal decision making are used to attempt in-roads into this philosophical landscape. This essay explores and develops the relationship between Markov decision processes (MDPs) and morality. A discussion of how MDPs and reinforcement learning are related to biological and social rewards extends to the proposal that humanity is steered by these mechanisms. The suggested way forward for humanity is to scientifically evaluate economic and social systems that focus on aligning ethical values with social and economic rewards.
Toby Asher Lightheart