Case in point. If you are Wiki and wrong it just sucks to be you and ai moderators or "monitored by users" as you say (if ai is training on wiki - ur words) .. "getting less than optimal quality". Yeah, that is fk'd and circular and just makes no sense - so, in conclusion - let's leave the arbiter of "good" information to be based on a LLM that consolidates the history of humanity's knowledge into a reply that takes about three seconds.