Re: Understanding Nanotech [PLEASE READ] [was Re: New government?]

hal@finney.org
Thu, 26 Aug 1999 10:24:09 -0700

Robert's essay was very interesting and enlightening. I haven't seen this kind of detailed analysis before. But 5 years to build a Gates-style mansion? I had no idea it would take so long.

A couple of quibbles:

> Then you get your open-source nanoseed to assemble solar collectors
> over most of the property. At an insolation of ~1000 W/m^2 and
> 0.2 conversion efficiencies (pretty conservative), that gives you
> 400,000 watts of power during the day. Assuming a mass manipulation
> cost of ~15,900 kJ/mol of sapphire (perhaps the highest cost), that lets
> you nanoassemble ~10 kg of nanomaterial per hour.

I wonder if this energy cost isn't off by a couple of orders of magnitude. There is not much discussion in Nanosystems of the actual energy costs to build the products. It seems that every synthetic step which Drexler analyzes produces energy rather than consumes it, which doesn't seem right, but I think he is assuming high-energy feedstocks. If you are starting with oxidized carbon, aluminum and silicon (like you'd find in the air and in the soil) then you first have to expend energy to get these things ready to use.

But if I do try to use the figures for energy dissipation on p441, I find that we should be able to construct 1.1 kg of product in one hour dissipating 1.1 kW. This is about 3.6 x 10^6 Joules of energy dissipated per kg. However on p397 Drexler says that it should be possible to dissipate only about 1.5 x 10^6 Joules per kg. Not a big deal, but a factor of >2 inconsistency is puzzling, when Drexler is giving us 2 significant figures.

In terms of sapphire, Al2O3, molecular weight about 100, a mole would be about 100 grams, so 1 mole would be about .1 kg. We should be able to construct a mole for .1 the kg cost, or 1.5 x 10^5 J if we use Drexler's lower estimate. Robert has 1.5 x 10^7 which is two orders of magnitude higher.

I want that factor of 100 back! Then I can build my mansion in 20 days rather than 5 years.

> It turns out that you probably don't want to assemble more than
> 10 kg/hour because if everyone on the planet is doing it you
> start to interfere with the heat carrying capacity of the planet [3].

How would you interfere with the heat carrying capacity if you're getting all your energy from sunlight? Isn't that virtually all dissipated as heat now anyway, minus a tiny fraction captured chemically, in both biological and nonbiological processes? It would seem that all we could hope to do is to capture it more efficiently and bind more of it chemically, so that if anything there would be less heat radiated. I don't see how we could make the earth hotter than it is already, just working with the already-present solar energy.

Or is that the issue? By "interfere with the heat carrying capacity" is it meant that we might make the earth too *cold*? (Sucking CO2 out of the atmosphere to get carbon for construction purposes might further cool things off.) That ought to be easy to fix, by sending up a few solar collecting satellites.

The total solar energy impinging on earth is about 2 x 10^17 Watts, which with a population of 10^10 gives everyone over 10^4 kW to play with. That would be enough to assemble 10^4 kg/hour, four orders of magnitude more, and lets us build our mansion in 4 hours instead of 5 years. However this would require an infrastructure to capture solar energy over much of the globe and pump it where needed.

Hal