Content moderators who worked on ChatGPT say they were traumatized by reviewing graphic content: 'It has destroyed me completely.'

L4sBot@lemmy.worldmod to Technology@lemmy.world – 225 points – Content moderators who worked on ChatGPT say they were traumatized by reviewing graphic content: 'It has destroyed me completely.'
businessinsider.com

Content moderators who worked on ChatGPT say they were traumatized by reviewing graphic content: 'It has destroyed me completely.'::Moderators told The Guardian that the content they reviewed depicted graphic scenes of violence, child abuse, bestiality, murder, and sexual abuse.

37

You are viewing a single comment
View all comments

It's the same story with pretty much any platform. While you may know YouTube as the tool where AI/software randomly bans innocent channels and demonetizes videos based on false positives, it actually has a small army of human moderators. These poor folks are constantly exposed to some of the most vile content humankind produces >! ::: beheadings, rape, torture, child abuse::: !< etc.

I once worked on a project aiming to help their mental well-being, but honestly, while a step in the right direction, I don't think it made much difference.

Edit: attempting to nest multiple spoiler formatting

>!beheadings, rape, torture, child abuse!&lt;

That's not how spoiler markup works here.

::: spoiler It works this way instead. yo yo dirty stuff here :::

Edited to add: Apparently there's not exactly consensus across different interfaces for what spoiler markup is supposed to be. Aaaargh!!

Neither of those appear formatted as spoilers in Voyager (formerly WefWef).

Edit: But, appears properly in Memmy.

Oh interesting. I just used the spoiler button while paying via Sync.

Hmm also in sync and nothing got spoiler censored