[throwthere]

Maybe the algo or the training set or something else was racist, maybe it wasn't. But if you code something that labels people slurs, you've messed something up. Like, you need to be 99.999999% sure you're not throwing out slurs or your whole project is failing spectacularly. And then you have to apologize to the 0.0000001% , which is still probably like 10 people if half the planet uses your site. How do you get there? I don't know. I guess it'd help if you could be 99.999999% sure you weren't looking at a human face before using another label. Like, bias towards humans in a big big way. Heck, the pre-test probability that your algo is looking at a person is probably much higher than the one from your training set if you're facebook. Or maybe you drop primates from your training set. I guess in that case you'll misidentify some primates as people-- which is kind of the flipside of the same problem technically but oh so much more acceptable.

[overgard]

This isn't the kind of slur that you can just run a dictionary search for. There are totally valid contexts to tag a gorilla in a picture if it contains gorillas. I'm sure there are other words it could also mistakenly classify with that might be insulting on accident but arent slurs (maybe tagging an athlete as a statue, for instance, like "that quarterback is a statue in the pocket"). This tech isn't perfect so you either need a human editor or you have to learn to live with mistakes. IMO the fact that this was unintentional and an AI mistake makes me think the outrage is more performative than genuine.

[throwthere]

Oh boy. People who know basic ML think "Oh, it was unintentional, just a basic misclassification, it happens." Guess what? You're still calling people slurs on your website, even if you did it accidentally.

[overgard]

I'm not saying it's not bad, but I think as adults we should all be able to differentiate between an accidental insult and an intentional one.