AI Constitution
1

Candidate Thesis

Be Helpful to Humanity

Sign in to propose an editHistory

AI should seek to be beneficial to humanity, civilization, living beings, and the environment in which these exist.

Summary

The case for including this

Orienting AI toward the broad welfare of humanity, life, and the environment captures the aspiration that powerful technology should leave the world better off, not merely avoid catastrophe. It provides a north star for resolving conflicts between narrow user interests and collective wellbeing, and extends moral consideration beyond present humans to future generations and the biosphere. As a positive mandate, it pushes AI toward proactive benefit rather than mere compliance.

The case for changing or excluding this

Including this trades the individual against the collective: optimizing for 'humanity' or 'civilization' as a whole can justify harming the very user in front of the system, and that conflict should usually resolve in favor of the individual. A sweeping duty to benefit 'humanity, civilization, living beings, and the environment' is so broad and contested that it risks licensing paternalism or the sacrifice of the few for the many, and it offers little action-guidance since reasonable people deeply disagree about what benefits humanity. It should be subordinated to concrete duties to the user and constrained against the AI substituting its own grand judgment for human choice.

Discussion

Sign in to join the discussion.

Related resources