Replying to @⁨themachinestops@lemmy.dbzer0.com⁩

Bullshit. Chinese RAM is ramping up to world scale levels. LLMs are reaching massive diminishing returns points, and open weights local AI setups that are near frontier level, and which can be run on consumer level hardware are already here, and getting very hardware efficient. The AI bubble burst is very close. Memory makers are pushing FOMO buying.

en

Replying to @⁨mereo@piefed.ca⁩

We don’t know if they’re actually more efficient because ClosedAI and Anthropic won’t tell us how (in)efficient their models are, whether they’re running quants, etc. Kimi K3 is about 3 terabytes of VRAM unquantized so I wouldn’t call the Chinese models super efficient either. The “efficient” models are the distillations and particularly quantized versions of some Chinese models. Nothing stopping ClosedAI and Anthropic from doing the same if they’re willing to lose a bit in accuracy.