Re:Kenyan moderators decry toll of training of AI
by xboxnolifes on Thursday August 03, 2023 @03:08PM
(#36990319)
One of them seems to have actually traumatic situations.
[ Parent
| on HN ]
Re:Kenyan moderators decry toll of training of AI
by giraffe_lady on Thursday August 03, 2023 @03:42PM
(#36990761)
The potential harms of reviewing fucked up material even fictional are well attested. It's factored into law enforcement & child porn investigations and has been for years.
I am assuming you knew that though, and just enjoyed this little moment of belittling someone for their suffering. Hope it was nice. Hope you're having the day you wanted to have.
Re:Kenyan moderators decry toll of training of AI
by xboxnolifes on Friday August 04, 2023 @02:44AM
(#36996280)
I really don't know it, because I truly cannot imagine how text could cause trauma to a normal person in the position of a third-party. And I'm assuming normal here, because they decided to do it as a job.
Re:Kenyan moderators decry toll of training of AI
by defrost on Friday August 04, 2023 @02:54AM
(#36996358)
They had no advance warning of what to expect:
The moderators say they weren’t adequately warned about the brutality of some of the text and images they would be tasked with reviewing, and were offered no or inadequate psychological support.
As noted above and below, it wasn't just text: The 51 moderators in Nairobi working on Sama’s OpenAI account were tasked with reviewing texts, and some images, many depicting graphic scenes of violence, self-harm, murder, rape, necrophilia, child abuse, bestiality and incest, the petitioners say.
> because I truly cannot imagineYou're not alone, there's likely a higher than normal percentage of HN'rs that have difficulty imagining how other people react and feel.