Can collective intelligence solve alignment?

Could a system that's designed to reward the evaluation of propositions be used to align ASI?

Determining the beliefs of 8 billion free individuals is required to build an AI that's aligned to all of them.