Model Better Governance
Our current form of government was written by candlelight; during a time when getting on a horse and yelling "the British are coming!" could change the course of history.
A lot has changed since then.
Our world is more complicated, connected, and concentrated. The rate of major societal change is faster than in the days of our founders, and it's getting faster.
This raises questions:
Is our current form of government perfect for the present realities?
Can it be better?
How is "better" defined, measured, and implemented?
To help design a better government by providing mechanisms to:
Identify factors affecting governance and how they interact
Encourage robust and precise discussions
Provide mechanisms to:
Create software models of governance proposals
Allow AI and human players to play the system; creating data
Improve the models with suggestions from human players and data analysts
Let anyone analyze the data produced by a game, make suggestions, or fork the model to test theories themselves
All of the core technologies already exist:
Creating scalable, human-understandable models
Creating AI that play games better than humans
Big data analytics
Systems for accepting and discussing feature requests
Putting it all together will take a lot of work, but the building blocks are there.
A software model is a mechanism for precisely describing a governance proposal and exploring its implications. Unlike an English description of a proposal, code is precise. Unlike a precise mathematical description, code can be interactive.
Precision and interactivity are important. The real world is complex. Correlation and causation are not always distinguishable. By precisely defining and varying a model, correlation and causation can be explored.
If an analyst believes that they have identified a causal factor, they can make changes to explore what happens. Likewise, if an analyst disputes a causal relationship, then they can make changes to see what happens.
You can think of a model as a video game. A game has rules and players. If someone likes a game, but wants to change it a little, then they can "mod" it. "Modding" means that they can change the parameters or the rules in a way that wasn't part of the original game creator's design.
Varying models and analyzing the results allows a robust and precise discussion of both the models and their implications.
Players: Human and AI
Players explore the models by playing the game. The resulting data is available for analysis.
AI players explore the game as it is written. The data that they produce provides a "flight envelope" of the potential outcomes of a model.
Human players add creativity and the "what if" factor for extending the game. It is the nature of human players to:
Want a more immersive game
Request new features so that they can try new strategies
Exploit or complain about unbalanced systems
These feature requests and exploitative behaviors can be used to improve the model; just like a normal video game.