2 ms·
I have every reason to assume the traumatising content examples are the point of the data classification exercise — I mean, filtering such content is probably t
by ben_w 3y ago
I have every reason to assume the traumatising content examples are the point of the data classification exercise — I mean, filtering such content is probably the single largest expense of large forums and social media sites, so of course loads of people would want to train an AI to do it for them.
As for giving people agency over their working conditions… I don't know — every so often there's an argument about prostitution, where people argue against it because it's dangerous, and then someone points out how all work carries some risk and someone says something about capitalism forcing men to effectively sell their lungs to coal mines which are much more dangerous than women selling their bodies, and everyone gets cross and a few months later there's a similar debate like nobody learned anything.
- brightlancer 3y ago> and everyone gets cross and a few months later there's a similar debate like nobody learned anything. This sums up online discussions for 30+ years. While some folks never learn, there's also a steady stream of new folks who weren't part of the previous conversations and _will_ learn if given the opportunity. (This is the positive side of the Eternal September.)