_rpd comments on The iVAIS Manifesto: Safety Through Character, Not Compliance

[_rpd]

Well argued, although I do think you can get a lot of safety from a constitutional AI rule like “in all thoughts and actions, be a good ally to humankind.”