Post

Imagine the FbI investigators using AI to catch the worst offenders. That would make them absolutely useless. But given the face of the FBI these days, they probably ARE using AI models for their investigations more than they should.

I block is not jail time. A ban is not jail time. Using an AI to trawl for CSAM then confirming with a human would not be a bad way to find investigations to start.

Who said anything about blocks being jail time or a ban being jail time?

Not to mention literal crime rings in other countries abusing and extorting workers Anyone that posts we need to go backwards on moderation needs to be literally duct taped to an IKEA chair to watch 12 hours of twitter, clockwork orange style, or lose their home

That’s a tad extreme.

I mean that's what we were doing to people in other countries. Seems fair.

There are a number of places that use an AI or even just an old fashioned Bayesian layer to filter the truly heinous, but have a small-scale human layer on top. Reddit and Mastodon spring to mind (YMMV on Mastodon's spam/heinous content-filtering)

people come back from wars less fucked up than content moderation was doing to people over time, yeah, especially if they were having to deal with significant volumes of CSAM

i think if you're honest with yourself just asking "how much would you have to pay me to think about the worst online interaction I've ever had for eight hours a day" would reveal that this is a dogshit job even before accounting for how much worse the internet gets than your worst interaction