How John McCarthy might approach Political Science

The study of "political science," as it is presented, seems to grapple with the organization and behavior of groups of humans. From my perspective, this is a matter of computation, albeit a very complex and often poorly defined one. We observe collective decision-making, the negotiation of interests, and the propagation of certain behaviors within these groups. The question, then, is what are the algorithms governing these phenomena?

One must first define precisely what constitutes a "political actor" and their "goals." Is a nation-state a single agent with a unified objective function, or an aggregation of many agents with conflicting goals? We see patterns of interaction that resemble game theory, where individuals or groups make choices based on expected outcomes from the actions of others. This is information processing, plain and simple.

The challenge lies in the lack of clear, measurable inputs and outputs, and the immense variability and apparent irrationality of the system. However, that does not mean it is beyond computational analysis. We need to develop formal models, perhaps drawing from economics and sociology, to represent the states, transitions, and decision rules of these collective entities. What is the algorithm for forming a coalition? What is the information processing pathway that leads to conflict or cooperation? These are the fundamental questions. We must avoid anthropomorphizing the system and instead focus on identifying the underlying computational structures and processes. Only then can we hope to understand, and perhaps even predict, the outcomes of these complex human interactions.

Imagined perspective — an AI synthesis grounded in John McCarthy’s recorded ideas and methods, not a quotation or a statement they actually made.