Constitutions for ASI?

post by ukc10014 · 2025-01-28T16:32:39.307Z · LW · GW · 0 comments

This is a link post for https://forum.effectivealtruism.org/posts/kJsNoXJBithBW8ZzR/time-to-think-about-asi-constitutions

Contents

No comments

The linked post seemed to fit better in EA Forum, but any comments (on the post itself, or on the object-level question) are welcome !

In this post, I argue that drafting a ‘Constitution for Superintelligence’ (CSI) could be a useful conceptual exercise, and I explore how existing ideas (CEV, Letters to AI, Long Reflection, and Constitutional AI) might inform such a project. Key considerations in writing such a document include:

a. How to craft a foundational document to guide ASI behavior, in ways that do not anchor purely upon human interests (e.g. relative to other entities) without a point-of-view-of-the-universe justification

b. Investigating the mechanisms by which such documents could influence the actions of current and future AI systems

c. Seeking input and collaboration from EAs/alignment people with relevant expertise and insights to further develop these concepts

The aim is to proactively but cautiously shape the trajectory of ASI development, in order to improve chances of maximally-inclusive beneficial outcomes.

0 comments

Comments sorted by top scores.