Morgenstern said:
No, you wouldn't need anti-matter (whole other topic which I'll leave to Pbartender

)
Simply put, anti-matter is thoroughly impractical as a power source or weapons-grade explosive. It's far too expensive and troublesome to make and to store. that sort of technology, at this point, doesn't even fall into the Hard Sci-fi category.
but usually when nono-tech gets presented in fiction you dump it on a pile of wood, and and up pops a chair after a (very) short while. As if the nanite were solar powered... or perhaps (very) convienently wood-powered. Nanites are damn small - that's the point, so they really can't be solar powered

- no surface area. Most likely they'll have to be chemically powered, so are a lot like bacteria.
Morgenstern said:
Exactly. Bio-engineered viruses and bacteria are nano-bots. And we already use them in that capacity. For example...
Converting alcohols into sugars.
Producing medicines, such as penicillin or insulin.
Cleaning up environmental wastes.
Synthesizing vitamins.
And other more recent and exotic things, such as...
Running a microscopic motor.
or
Salvaging precious metals from electronic components.
Morgenstern said:
But as that stuff is wandering around my body, I want it to show restraint on a body-wide scale. That means the nanites have to be superbly networked so they don't snatch sugars in areas that are highly sensitive - i.e. "no feeding in my brain please. Get lunch in my muscle tissue before commuting up there to work. Ok?" That mean more energy loss (in communication and navigation) and more size/complexity of the nanite (having a working understanding of location a typical body or coding guidelines/tables for decision trees based on sugar availability). That extra work makes me wonder if it really will be all that vastly more efficient than other tools available to us.
It's not as tough as you think, Morg...
The trick is to build your nanotech (GEed bacteria or virus) to do a specific job in a specific part of the body. Bacteria already do that sort of thing... A bacteria geared for giving you food poisoning has a very hard time living anywhere outside your stomach and intestines. And you can't get tetanus by eating a rusty nail. You engineer a trait into the bacteria so that it simply dies off after a certain amount of time... Or have certain biochemical triggers that activate the bacteria or render it "dormant". The presence or absence of a certain concentration of adrenaline, for example, would be a good trigger for fast-healing nanites.
Part of the problem is scale. The chair example is perfect for that - you are expecting machinery capable of moving atoms around to produce an object in the human scale - devices with an individual range of perception in the nanometer range are going to have to produce a object with smooth lines in the meter range - 1 atom at a time. In other words, getting it right 50 Quadrillion times along the axis (conservatively). It's like humans building a 10,000 mile long road in a perfectly straight line - by eyeballing it. If you want to give them survey equipment its going to take extra size to store that coding and every ditch digger doesn't need the full package.
Again, it's not quite as difficult as it seems.
We've already got engineered bacteria that can build a perfect crystaline lattice or polymer chains. And that's the key. Crystals get built in naturally straight lines. Nanites (in bacteria form) would be great at building perfectly regular shapes. All the nanite needs to know is "Find this kind of molecule and stack this molecule on top of that molecule in this way. Do it again." Soon, you've got a spool of thread, or a rod, or a flat plate, or a disc, or a sphere or whatever.
Also, nanotech would be great for replacing one type of material with another. That's an easy one.
Nano-tech IS real, and in its infancy. Electricity is alos real, and we tried to attribute ressurection to its many promises early on. We did the same thing with radiation, and we did it again with genetics. Nano-tech is the new magic
du jour of fiction, and its real world incarnation will get better and better as it matures, but a lot of early thinking on it is just hooey, same as every other revolutionary advance. Its not going to work as fast as people now think, be as clean as people think, be as cheap as people think, and it's going to be a lot harder to use inside bodies than most people think. And by people I mean many of the sci-fi writers who are giving us our perceptions of where this technological concept will be when it does mature. That's mostly where
I am going with this

.
All that's true, but I don't think certain nano-tech jobs will be quite as difficult as you think. Industrial and medical fields will be their strengths, since those are within the provices of a bacteria's natural environment, purpose and operating parameters.
Where people get screwed up is thinking that nano-tech will be actual "tiny robots", or that it will have any sort of computing/intelligence or communication powers. That's rediculous. Real world bacteria nanites will be specifically programmed to do a specific job under specific circumstances, and nothing else.