misk@sopuli.xyz to Technology@lemmy.worldEnglish · 9 months agoBaidu CEO warns AI is just an inevitable bubble — 99% of AI companies are at risk of failing when the bubble burstswww.tomshardware.comexternal-linkmessage-square194linkfedilinkarrow-up1740arrow-down122
arrow-up1718arrow-down1external-linkBaidu CEO warns AI is just an inevitable bubble — 99% of AI companies are at risk of failing when the bubble burstswww.tomshardware.commisk@sopuli.xyz to Technology@lemmy.worldEnglish · 9 months agomessage-square194linkfedilink
minus-squarePennomi@lemmy.worldlinkfedilinkEnglisharrow-up17·9 months agoIf bitnet takes off, that’s very good news for everyone. The problem isn’t AI, it’s AI that’s so intensive to host that only corporations with big datacenters can do it.
minus-squareStarbuncle@lemmy.calinkfedilinkEnglisharrow-up17·9 months agohttps://www.microsoft.com/en-us/research/publication/bitnet-scaling-1-bit-transformers-for-large-language-models/ use 1 bit instead of 8 or 16, yay performance gainz
minus-squarebrucethemoose@lemmy.worldlinkfedilinkEnglisharrow-up1·edit-29 months agoWhat star said, but what it also does is turn hard matrix multiplication into simple addition. Basically, AI will be hilariously easy to run compared to now once ASICs start coming out, thought it will run on CPUs/GPUs just fine.
If bitnet takes off, that’s very good news for everyone.
The problem isn’t AI, it’s AI that’s so intensive to host that only corporations with big datacenters can do it.
The fuck is bitnet
https://www.microsoft.com/en-us/research/publication/bitnet-scaling-1-bit-transformers-for-large-language-models/ use 1 bit instead of 8 or 16, yay performance gainz
What star said, but what it also does is turn hard matrix multiplication into simple addition.
Basically, AI will be hilariously easy to run compared to now once ASICs start coming out, thought it will run on CPUs/GPUs just fine.