Ethical Regulators and Super-Ethical Systems


I believe that my paper “Ethical Regulators and Super-Ethical Systems” will be of interest to some of you.


The paper builds upon the Good Regulator Theorem and the Law of Requisite Variety to address the issue of making systems ethically adequate. The resulting Ethical Regulator Theorem identifies nine requisites that are necessary and sufficient for a cybernetic regulator to be effective and ethical.

The paper also proposes a new classification framework for cybernetic systems, which accommodates 1st and 2nd order cybernetics, ethical systems, and superintelligent systems. That framework highlights the impending bifurcation into either a technological utopia or a technological dystopia, for which I propose concrete actions to help steer humanity towards the utopia and away from the dystopia.

Finally, it pushes some conventionality boundaries with a spiritually empathic call-to-action.

You can get the PDF from:

I would appreciate any constructive feedback or questions.

Thanks in advance,



(I’m still reading it.)

Certainly a most important issue that’s not solvable be simply FUD’ing about the existential risks of AI.

Partnership on AI: Thematic Pillars

DeepMind Ethics & Society


From :

Can one write an algorithm for Kant’s moral calculus? Is there anyone who has tried?


A few categorical links:


This was a relevant and interesting talk:

“AGI 2011: The Future of AGI Workshop Part 1 - Ethics of Advanced AGI”