Yes, and automation is very difficult to use for these tasks. Running such huge infrastructure can benefit in many ways by automation, just look at netflix. Bu

Yes, and automation is very difficult to use for these tasks. Running such huge infrastructure can benefit in many ways by automation, just look at netflix. But it doesn't function well in these cases, and again and again it's clear these companies can't employ enough people and tools to prevent blantly illegal content (to the point many of those hired to review potential illegal material come away from that job with PTSD) much less identify and moderate hate speech or even speech that is more nuanced. Reddit did some stuff right with sub moderators, but even then its a whole charlie-foxtrot when the moderators encourage it or aren't active (and thats excluding ones with their own specific agendas). At this scale, these companies cannot moderate their platforms properly, and automation could help, but isn't the solution until ot greatly improves.