They get free training data from users to make the models better. Worse, it seems like content that slips past human 'AI slop' detection gets rewarded more, whi

They get free training data from users to make the models better. Worse, it seems like content that slips past human 'AI slop' detection gets rewarded more, which just pushes people to write more convincingly fake stuff.