Clock speeds are not and have not been the limiting factor in social impacts of computing since the 90s. The limiting factor is software architectures, and org

Clock speeds are not and have not been the limiting factor in social impacts of computing since the 90s. The limiting factor is software architectures, and organizational architectures, which are extremely primitive. Consider the basic notion of contractors fulfilling simple tasks, a la Lyft/Uber. That has only been around for a few years and we can only do it for the simplest and most unambiguously evaluable tasks like “drive to here and then here”. Also consider that the absolute simplest programming tasks, like “make a parameterless function that calls several other parameterless functions”. This is a task any literate person could do and yet it’s technically out of reach of 99% of computer users. Not because of clock speeds, but because we don’t understand software architectures or human tool use well enough to design it. Those technologies will someday be widely available across a broad range of industries to all levels of employees. But we are not close to that time. There is an incredible amount of basic research required to accomplish it. And an utterly trivial amount of CPU resources.