Re: How not to start the day

On 05/12/2024 21:52, Justisaur wrote:

As to their algorithm, I believe you posted about them changing to AI moderation, which is orders of magnitude more difficult, or impossible to figure out what the hell it's doing in the black box of it's program.   From what I've read on it, basically no one knows what it's doing and all they can do is retrain it, or implement layers on top of it that try to stop it from doing things they don't want it to.  Then there's the fact they may not even actually try to stop it from doing things that they *say* they don't want it to, or that one would expect fairness out of.

My experience of algorithms, mainly FB, is that they are really bad at detecting sarcasm. So I once said 'burn the heretic' and I got hit with a seven day ban for extremism language. Do they not understand Brits have built a whole way of life around sarcasm.
In general I kinda agree with you that companies are more than happy to accept some users as collateral damage as long as they can avoid getting a reputation of allowing the likes of hate speech, well unless it's twitter, sorry X, which seems to embrace it with open arms.