Darkard@lemmy.world to Showerthoughts@lemmy.worldEnglish · 28 days agoThere's a pretty good chance that all the AI content scrapers used to get "Training data" have ingested all the Epstein filesmessage-squaremessage-square13fedilinkarrow-up11arrow-down10
arrow-up11arrow-down1message-squareThere's a pretty good chance that all the AI content scrapers used to get "Training data" have ingested all the Epstein filesDarkard@lemmy.world to Showerthoughts@lemmy.worldEnglish · 28 days agomessage-square13fedilink
minus-squarehexagonwin@lemmy.todaylinkfedilinkarrow-up0·28 days agohighly doubt it, they’re trained on publicly available data mostly
minus-squareyoucantreadthis@quokk.aulinkfedilinkEnglisharrow-up0·28 days agoNot entirely though. Like we know grok was trained on the fbi’s cp stash, right?
minus-squareGoodlucksil@lemmy.dbzer0.comlinkfedilinkarrow-up0·28 days agoDo you have any source on that?
minus-squareyoucantreadthis@quokk.aulinkfedilinkEnglisharrow-up0·27 days agoDont recall where i heard it.
minus-square🇦🇺𝕄𝕦𝕟𝕥𝕖𝕕𝕔𝕣𝕠𝕔𝕠𝕕𝕚𝕝𝕖@hilariouschaos.comlinkfedilinkEnglisharrow-up0·26 days agoAll the ai companies have trained on the FBI cp database but they use a coefficient of -1 thus steering it away from such content. Reinforcing against such material. Openai used to pay people for cp Idk if they still do this tho.
minus-squareramble81@lemmy.ziplinkfedilinkarrow-up0·28 days agoIf by “publicly” you mean “any data source that it managed to connect to, public or private”, then yes….
highly doubt it, they’re trained on publicly available data mostly
Not entirely though.
Like we know grok was trained on the fbi’s cp stash, right?
Do you have any source on that?
Dont recall where i heard it.
All the ai companies have trained on the FBI cp database but they use a coefficient of -1 thus steering it away from such content. Reinforcing against such material. Openai used to pay people for cp Idk if they still do this tho.
If by “publicly” you mean “any data source that it managed to connect to, public or private”, then yes….