3 ms·
I think a reason those 2-3 historical cases are not an eternal motif is that humanity is probably approaching the feasible limit of Type I on the Kardashev scal
by i_am_proteus 8y ago
I think a reason those 2-3 historical cases are not an eternal motif is that humanity is probably approaching the feasible limit of Type I on the Kardashev scale.[0] We're at the first point in human history where reducing aggregate energy consumption is viewed a serious goal by most (a lot?) of the fully-industrialized world.
Previous technological advances have caused individual output to rise because individual humans were able to harness more energy. A human with a tractor does more work and consumes more energy than a human without a tractor.
Some new technology is letting us do more work with less energy input, but a lot of new technology is just automating the human out of the process and using just as much energy to do the work. One person can supervise multiple CNC mills, but those mills' power draw isn't much different than old manual Bridgeports (hell, they might actually be old manual Bridgeports with a few mods).
[0]https://en.wikipedia.org/wiki/Kardashev_scale https://en.wikipedia.org/wiki/Kardashev_scale
- Atheros 8y agoWhat are you talking about. We use a minuscule amount of the sun's Earth-bound energy to make electricity and have so much cheap fossil fuel that we haven't even started on fusion reactors yet.
- i_am_proteus 8y agoMy conjecture is that "using all of the energy" is not going to be the limit. Perhaps I am using the Kardashev scale too loosely. Unlike previous industrial revolutions, the current revolution in automation has not seen economic growth result from doing more work by unleashing more energy. Economic growth has generally been a result of using existing energy sources more efficiently but also from reducing labor costs.