Can collective intelligence solve alignment?

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

If Doug Lenat's work would have scaled, Eliezer Yudkowsky would not have existential concerns about it.