If we can't even solve the human/human alignment problem at scale, why do we think we can solve the human/AI alignment problem? Maybe they're the same problem and maybe the solution is a monetary standard that is bounded to the same constraints as all the agents trying to scale.