Not the type of "moderation" that YouTube currently engages in, which is trying to enforce some undefined moral standards over content. I meant the sort of ranking that web search does, where the value function is far more precise: the best results are whatever the user was looking for, regardless of why they were looking for it. The moderation can be automated because it's just down-ranking content that's trying to manipulate the algorithms. Then if people do searches for bad things you just stay silent - it's a problem between that user and their society/legal system if they're searching for bad things, not between the user and their search engine.

This system has worked well for a long time. You do sometimes see stories like "murderer caught after police discover he was searching for how to hide bodies" and things like that, but nobody asserts the murder was Google's fault, no more than they would a library.

Gmail works the same way. Spam is whatever Gmail users say it is, via their "Report spam" button. It's a form of collective voting, not a value judgement by a few executives. Chrome blacklists websites when they're hosting phishing or malware, but that's because those sites interfere with the correct operation of the infrastructure itself. It's not because some random L4 SWE threw a hissy fit.

The sort of moderation YouTube is doing is inviting trouble because it's not so much a slippery slope as a vertical cliff-edge. Once you bow to requests to manually purge "bad" content, you end up in an endless loop in which there are always more requests for purges and the demands to do so get increasingly vicious. A strong and total stand of the form "we are merely dumb pipes" is the most effective way to stop a never ending purity spiral, and it's one some other products have done successfully: nobody ever asks Chrome to nuke websites even though they could, because everyone knows the answer will be a furious rejection of the request (at least up until now), so they don't even bother asking.