stopthatgirl7 to Technology@lemmy.world • 2 years agoEverything Made By an AI Is In the Public Domaindoctorow.medium.commessage-square137fedilinkarrow-up1650arrow-down130file-text
arrow-up1620arrow-down1external-linkEverything Made By an AI Is In the Public Domaindoctorow.medium.comstopthatgirl7 to Technology@lemmy.world • 2 years agomessage-square137fedilinkfile-text
minus-square@illusoryMechanist@lemmy.worldlinkfedilinkEnglish2•2 years agoDoes this mean we can, indeed, train AI on ChatGPT/other AI outputs?
minus-squareNatanaellinkfedilinkEnglish8•2 years agoFYI there’s studies showing that repeatedly training ML on ML outputs degrades quality sharply after 4-5 cycles
minus-square@Prager_U@lemmy.worldlinkfedilinkEnglish0•2 years agoThat is a used method for training other LLMs, I believe OpenAssistant did this.
minus-square@assassin_aragorn@lemmy.worldlinkfedilinkEnglish0•2 years agoDoesn’t that break the models?
minus-squareNatanaellinkfedilinkEnglish2•2 years agoIt does after 4-5 cycles yeah, built in algorithmic biases gets amplified to a degree that it breaks the model
Does this mean we can, indeed, train AI on ChatGPT/other AI outputs?
FYI there’s studies showing that repeatedly training ML on ML outputs degrades quality sharply after 4-5 cycles
What does ml stand for
Machine Learning in this context
That is a used method for training other LLMs, I believe OpenAssistant did this.
Doesn’t that break the models?
It does after 4-5 cycles yeah, built in algorithmic biases gets amplified to a degree that it breaks the model