How to express this system for ethically aligned AGI as a Mathematical formula?

post by Oliver Siegel (oliver-siegel) · 2023-04-19T20:13:43.881Z · LW · GW · 0 comments

Contents

  Background:
  Framework for ethical solutions:
None
No comments

I have a graph-based solution for decision-making in situations with multiple agents and stakeholders. I wonder if there's an elegant way to express it as a Mathematical formula.

Background:

I have oriented myself on the "true justified beliefs" framework from epistemology. Here, propositions are categorized into beliefs, and those beliefs that are properly justified are considered knowledge.

Framework for ethical solutions:

According to my proposed framework, every proposition or proposal can be classified as a problem, a terminal goal, or an instrumental goal.

Once classified, further propositions can be generated, that justify or criticize the alignment of a solution.

Specifically, solutions can be validated by listing all the problems they solve and all the goals they fulfill. Solutions can be criticized by listing all the negative consequences that arise from them.

This should give us an "alignment ranking" among all submitted and properly categorized proposals.

I'm not very good at Math, especially not at Mathematical notation. 

I know that the illustration of sets above can be written as a vector matrix or as a node graph, but I don't know how to do this. 

Who can help?

0 comments

Comments sorted by top scores.