From a comment yesterday: "We will see nanotech-like things happening, but not the build-anything just-wish-for-it nanofactory. We'll have new kinds of materials, .... But the pure build-from-the-atom-up nanofactory will never exist. It's too specific a vision, and those kinds of visions never come true."
It's certainly true that flying cars have been few and far between, and the closest thing we have to the Jetsons' robot maid is a Roomba. Is a nanofactory a similarly specific and unlikely vision? Are there practical problems lurking that will make it undesirable even after it becomes possible? After all, flying cars have existed for a decade or so, and helicopters for a lot longer than that. We could have had a VTOL aircraft in every garage if we'd wanted it.
We can break this question down. First, will nanofactory technology be developed? Second, if it is developed, will it be widely deployed?
The basic purpose of a nanofactory--building complete products rapidly from simple feedstock--is being pursued today, by a number of groups, in the form of rapid prototyping (RP) technologies. Today's RP cannot build products--it can barely integrate wires, and has a lot of trouble with motors, bearings, and so on. But these may come in time.
So what's the difference between a futuristic RP machine, capable of building complete products, and a nanofactory? The answer: The products of a nanofactory would have many orders of magnitude higher performance. And there's no known way to get that kind of performance, other than building machines from the atoms up. And there's no known way to get large-scale products, other than exponential manufacturing. And a nanofactory looks like a very efficient way of doing exponential manufacturing.
Assuming nanofactories are developed, will they be used as home appliances? It's a good question. It would have been easy to assume, in the 1950's, that every teenager would want and obtain a helicopter, but a combination of cost, safety, learning curve, bureaucracy, and rotor noise made things turn out differently. In the final analysis, a helicopter is only marginally more useful than a car for most family purposes--and useless for cruising down Main Street.
Let's consider a different kind of technology--electricity. It is powerful, convenient, and has many applications that nothing else can replace. Once the appliance-sized electric motor was invented, there was no question but that everyone would want electricity. (Electric lights were only marginally better than gaslights. But nothing else could put a significant fraction of a horsepower in a handheld device.)
So, will a home nanofactory do a wide range of useful things that no competing technology can do? It should be able to build a huge range of products, direct from blueprint, at low cost. Will that be more than incrementally better than going to the store or service bureau, and using their nanofactory? This is a good question.
What will be the downsides of personal nanofactories? Bureaucracy is likely to be one: intellectual property and public security may require onerous licensing or other restrictions. If the software is poorly designed, security for the user might be a problem--you don't want spam climbing out into your house. The first color laser printers--two-dimensional general-purpose reproduction machines--were the size of a large desk, required regular expert service, and smelled funny. The service bureau was the right place for them, even if they hadn't cost $100,000.
It's worth mentioning here the "nanoblock" idea. Instead of building nanomachines atoms-up and forming them into products in the same factory, you build a set of nanoblocks that is collectively general-purpose. The vast majority of the energy and time required to build a product is used to build the nanoblocks, not fasten them together. So for a modest reduction in product performance and flexibility, you get the ability to build a product in seconds rather than minutes or hours, and possibly better security (both for the user, and from the user).
So, instead of a nanofactory in every home, we might see a nanoblock-assembler using prebuilt nanoblocks made in a few central locations and shipped in cartridges. Of course, the markup on these would be at least as high as inkjet cartridges.
Note that while answering a skeptical question, we have worked around to something that, from the user's point of view, is even more advanced. So advanced that no other technology could compete (with the possible exception of other molecular manufacturing based technologies). So generally useful that everyone will want one--assuming the marketing department is even minimally competent.
It is not guaranteed, of course, that things will turn out this way. It is possible, for example, that virtual reality (possibly augmented by special-purpose hardware) will turn us into such couch potatoes that we won't have much use for physical products anymore. Arguments on this could go back and forth for years, and will only be definitively answered when we see how things actually turn out.