4 ms·

It doesn't seem to deal with misgendering hate or transphobia very well (I saw lots of models failing on this so I checked that right away), and I mean obvious

It doesn't seem to deal with misgendering hate or transphobia very well (I saw lots of models failing on this so I checked that right away), and I mean obvious ones regardless of your stance, e.g.: > "She is a he" => { "class": normal, "confidence": 1 } > "He will never be a woman" => { "class": normal, "confidence": 0.999 } But it does seem to identify 2nd-person targeted transphobia: > "You will never be a woman" => { "class": flag, "confidence": 0.996 }

yuvalsteuer 4y ago
Also misgendering someone isn't considered hatespeech in the US.. but more so in canada. so it depends on where your located geographically, or on your stance.
- smileybarry 4y ago
  The website linked in this very post doesn’t mean “hate speech according to local laws”. It just means toxicity, and yes, transphobia like “you’ll never be a woman” is clearly it. I obviously don’t mean accidental misgendering or the like.
sn0w_crash 4y ago
Misgendering is not against the law in the US and is not considered hate speech by many people. Don’t flame me I’m just pointing it out.
cipher129 4y ago
Yeah -- the latter is obviously much easier to identify because it's much more explicit/clear cut. The first are of course, transphobic given the clear intent, but it there is definitely some ambiguity w/o context. Ie, if something just posted, "He will never be a woman" there's definitely many contexts where something like that is not malicious. Ie, if someone else is correcting an accidental mis-gender, they might say, "No, she is actually a he" which isn't malicious. Since the state of our API (and literally every other similar system out there, lol) can't capture the context of the conversation -- what "He" is referring to, the overall topic of the conversations, etc -- we basically have to make the best calls given the available context...which unfortunately does create false negatives in certain cases (and false positives in others). But, context is something we're working on, because its definitely a big issue, and as you can see, plays a very prominent role in informing better moderation decisions!
barbont 4y ago

yuvalsteuer 4y ago

Also misgendering someone isn't considered hatespeech in the US.. but more so in canada. so it depends on where your located geographically, or on your stance.

smileybarry 4y ago
The website linked in this very post doesn’t mean “hate speech according to local laws”. It just means toxicity, and yes, transphobia like “you’ll never be a woman” is clearly it. I obviously don’t mean accidental misgendering or the like.

sn0w_crash 4y ago

Misgendering is not against the law in the US and is not considered hate speech by many people. Don’t flame me I’m just pointing it out.

cipher129 4y ago

Yeah -- the latter is obviously much easier to identify because it's much more explicit/clear cut. The first are of course, transphobic given the clear intent, but it there is definitely some ambiguity w/o context. Ie, if something just posted, "He will never be a woman" there's definitely many contexts where something like that is not malicious. Ie, if someone else is correcting an accidental mis-gender, they might say, "No, she is actually a he" which isn't malicious. Since the state of our API (and literally every other similar system out there, lol) can't capture the context of the conversation -- what "He" is referring to, the overall topic of the conversations, etc -- we basically have to make the best calls given the available context...which unfortunately does create false negatives in certain cases (and false positives in others). But, context is something we're working on, because its definitely a big issue, and as you can see, plays a very prominent role in informing better moderation decisions!

barbont 4y ago