Price/performance of computation from 1939 to 2023

[John Clark]

I found this chart to be very interesting, the plot is logarithmic therefore the trend is exponential. I think it explains why we didn't have large language models that were worth a damn in 1939, or even in 2020, but we do today. And there's no indication this exponential trend is abating.

John K Clark    See what's on my new list at  Extropolis

3cq