Discussion about this post

User's avatar
Ivan Vendrov's avatar

> arrangements like capitalism and democracy channel conflicts into formats that are productive (like business competition and political campaigning) rather than destructive (like theft or violence)

This distinction between productive and destructive conflict seems more normatively important than either coherence or well-foundedness; would love to see it expanded on.

Successful coalitional agents are defined more by their ability to channel internal conflict constructively than by being particularly well-founded or coherent: e.g. the Roman Republic, UK, US, Elon.

Expand full comment
Melon Usk - e/uto's avatar

Great article, Richard! Yep, the whole ethics can be expressed with a binary trees as the minimal model. Infinitesimal agents can do nothing, grow left, right or both ways. This way they make one of 4 choices.

They interact in one of 3 ways: give freedom (give space to grow to another/let another choose/help another), do nothing or take freedom (force something on another). We are writing this computation ethics paper to solve alignment from the bottom up.

45 pages with ~20 graphs done already, I can send you the draft, we need co-authors!)

By the way, we recently solved AI alignment with the top down approach: https://open.substack.com/pub/melonusk/p/notes-on-euto-principles-and-tenets?r=5d055t&utm_medium=ios

Expand full comment
2 more comments...

No posts