I know this is decentralized, but is there any way to moderate or censor offensive ratings? Or is that impossible cause of decentralization?
Its not bad if there is no censor, offensive ratings are acceptable as long as theyre not fake and backed by facts.
What about racism, sexism, that kind of stuff? Can the algorithm block that? Actually how do we know if the algorithm itself doesn't have a preference or distaste built-in?
Well I assue they could program something to find such stuff, but probably a human has to check and decide in the last instance?
Yes of course, the AI should be controllable by human, so I think there will be manual moderation.
But I don't see racism, sexism or any other things like that will be a good quality review or even backed by facts/truth and the AI will surely make it has no impact to the reputation.