Replacing Laws With AI – Basic Principles

If you wanted to replace laws with something else, what would that something be?

By laws here I mean the written statutes, the police who enforce, and the judicial system in general. This feedback system, built on the written word of these laws, is currently the basis of our civil society. It’s bones if you will.

But there are issues. Sociopaths don’t care about the laws, only lawyers actually read the laws, and the wealthy who can afford lawyers clearly are afforded preferential treatment. And so on.

So, let’s say we wanted to be true revolutionaries and start a society without laws? How would that work?

Well, since we are in thought experiment mode here, let’s assert that we can posit technologies that do not exactly exist yet. Here’s my swing at it:

Instead of idea of good, evil, and punishment, I would start with the idea that there are patterns of behavior that benefit people, and patterns that harm them. These patterns are made up of the actions of individuals, but also their internal states and the state of the economy and world. The benefit and harm needs to be considered in much longer time scales than human consciousness operates in.

These patterns of benefit/harm are far, far too complicated to be extracted and deduced in real time by humans. But a sufficiently advanced machine learning system could do a good job. Or maybe a benefital one.

Of course, this would have to start with you as an individual. The Singleton Pattern is the idea that people as individuals are what is important in the world. And an AI that was declaring something was a law just because it mysteriously matched some pattern but does not benefit you is about as dystopian as it comes.

So, let’s assert that each human has an AI looking out for it. And moment to moment, your AI is determining how your actions match a positive or negative pattern in relation to you and your relationships.

Then, there would be another level of AI that examines the output of these AIs to help humans guide policy of the society and species. Which in all likelihood will mean telling corporate AIs what laws they have to follow.

If we want to replace laws with something better, it has to start with AIs that defend us as individual entities. Otherwise, prepare to be rolled.

Pic from somewhere on Reddit.