Billy Brown, <ewbrownv@mindspring.com>, writes:
> Robert J. Bradbury wrote:
> > However, the calculation assumes that every person on the planet can
> > find *some* use for his 10 kg allocation of robots consuming ~100KW
> > of power. Right now most of the people on the planet don't come
> > close to consuming 100KW. A standard house (with 200 Amp service,
> > probably above average), only pulls 20KW when it is maxed out.
>
> Something is seriously wrong with these numbers. My handy reference lists
> the solar constant at 1340 W/m^2, and the mean radius of the Earth is
> 6.37x10^6 m. Multiply that out, divide by 3 billion people, and you get an
> approximate natural energy budget of 5.6x10^4 KW per person. We should be
> able to use whatever fraction of that we care to actually harvest without
> fear of creating a global heat problem.
Robert Bradbury kindly shared with me some material from Robert Freitas' upcoming book where he discusses these issues. It appears that the main concern is not so much the amount of energy being used, but the pattern and distribution of it. For safety, the ceiling is chosen so that industrial activities would not change heat dissipation patterns by a significant fraction of their natural values.
For example, if you captured all the solar energy but dissipated it proportionally to the current human population distribution, you would have extreme effects on global weather and possibly climate. The oceans would be cooled and the cities extremely hot. It might take years for these effects to be noticed but by then it may not be easy to fix them. Yes, with enough work we can probably deal with these problems, but the idea was to set a power-dissipation ceiling low enough that we could be pretty confident that the earth's natural heat regulation system would not be overloaded.
There is also the question of how efficiently we can capture this potential ~10^4 kW per person (and chances are the population will be 3 times your 3 billion figure in the nanotech era). Even with nanotech, conversion efficiencies will not be perfect, especially if we are transporting energy far from where it was collected.