Humans vs. Upload filters

A series of conversations today made me realize that one of the least expected outcomes of the implementation of Article 17 will be a massive increase in work for off-shore content moderators. While most of the discussion around Article 17 focusses on the fact that it will require platforms to implement automated content recognition tools to filter user uploads (the dreaded #uploadfilters), there is much less attention to the fact that along with these filters will come small armies of content human moderators to do the (now legally required) “human review” of the inevitable mistakes that the filters will make. I find the idea that, by talking about human review, the directive carves out a niche that is explicitly protected from takehover by AI oddly satisfying.