DamnianWayne@lemmy.world to World News@quokk.au · 22 hours agoTrump clears way for Nvidia to sell powerful AI chips to Chinawww.theguardian.comexternal-linkmessage-square16fedilinkarrow-up173
arrow-up173external-linkTrump clears way for Nvidia to sell powerful AI chips to Chinawww.theguardian.comDamnianWayne@lemmy.world to World News@quokk.au · 22 hours agomessage-square16fedilink
minus-squarenymnympseudonym@piefed.sociallinkfedilinkEnglisharrow-up1·17 hours agoNot ones to compete with NVidia Not ones that will train GPT5/6/7
minus-squarecassandrafatigue@lemmy.dbzer0.comlinkfedilinkarrow-up2·14 hours agoThe… That’s not Enough data to train a hypothetical gpt6 literally does not exist.
minus-squarenymnympseudonym@piefed.sociallinkfedilinkEnglisharrow-up1·13 hours agoMost frontier models since GPT4 actually use synthetic pre-training data , before slurping up the Internet and after along with RLHF
minus-squarecassandrafatigue@lemmy.dbzer0.comlinkfedilinkarrow-up2·13 hours agotoo many fingers sausage fingers
minus-squarenymnympseudonym@piefed.sociallinkfedilinkEnglisharrow-up1·13 hours agoAlso training data is totally independent of chip feature size, interconnects, memory bandwidth, and all the things actually relevant to chips
Lol what? They already did.
Not ones to compete with NVidia Not ones that will train GPT5/6/7
The… That’s not
Enough data to train a hypothetical gpt6 literally does not exist.
Most frontier models since GPT4 actually use synthetic pre-training data , before slurping up the Internet and after along with RLHF
too many fingerssausage fingersAlso training data is totally independent of chip feature size, interconnects, memory bandwidth, and all the things actually relevant to chips