In this post I pointed out that in the foreseeable future, nanotech devices are likely to be energy-starved. Chris Peterson asks in a comment whether there would be a problem from the heat dissipation from this energy use. The analysis is worth a post of its own, so here goes:
About 100 thousand terawatts of sunlight hit the Earth. About 50% of this is absorbed by the surface and becomes heat. That’s 50 thousand terawatts of heat.
World power usage is currently about 15 terawatts. So if we got all our power from ground-based solar conversion, we could increase energy usage by more than a factor of 3000 and not change the energy balance a bit. The energy going into the machines and ultimately coming out as heat would be the same energy that would have been directly heating the surface anyway.
Even at the rates of energy growth we saw in the 1960s, this won’t happen for another century. By that time, it is to be hoped, most of the heavy energy use (and, increasingly, the population) will have moved out into space where most (99.9999999%) of the solar power is to be found anyway.