Basement development? Big leaps?

by Eric Drexler on 2009/12/20

Part 2.1 of a series on the history and prospects of advanced nanotechnology
concepts, prompted by the upcoming 50th anniversary of Feynman’s historic talk,
“There’s Plenty of Room at the Bottom”.

A commenter on the previous post has aired the often-mentioned basement-breakthrough scenario for achieving high-level molecular manufacturing. This scenario makes no sense, and I probably should say why. One reason is generic, and another is rooted in chemistry and physics.

No basements

Before a small group with modest resources could accomplish anything dramatic, it will have become quite obvious (despite the cognitive bug I mentioned) that the implementation technologies have become very capable. In that time frame, those technologies will be exploited for a range of advanced but relatively mundane purposes by increasingly large and well-funded R&D programs worldwide. Because the implementation problems will involve complex devices and extensive laboratory facilities, they will be most effectively addressed by large teams of diverse and well-funded specialists. This is far, far from a basement scenario. (Slashdot readers take note.)

The basement-development idea is also incompatible with the modern trend toward broad, open, global collaborations. Even mathematical proof, the classic area for single-person breakthroughs, has recently succumbed to the power of open collaboration (the ‘Polymath Project’, recently discussed in Nature).

No leaps

The commenter also brings up another idea that should be dropped, the idea of a direct leap from accessible, starting-point technologies to the most advanced technologies that have been discussed. Because there is a myth that I have foolishly proposed this, I should explain why it, too, makes no sense.

It should be kept in mind that concepts for synthesis of diamond-like carbon-rich nanostructures involve using very rigid mechanisms to position highly reactive molecular groups in ultrahigh vacuum. Each of these properties is thoroughly incompatible with mechanisms built from moderately rigid biomolecular materials operating in aqueous environments. This incompatibility is so deep that some scientists (who evidently haven’t thought very hard about the problem or read my publications on implementation pathways, e.g., in Nanosystems) have had difficulty seeing how it is even possible to get from one to the other.

Accessible steps that make further steps accessible

The answer, of course, is that one doesn’t do so directly. It turns out, for example, that there are admirably stiff inorganic materials that crystallize spontaneously in water, and are natural targets for controlled synthesis by biomolecular systems. See, for example my posts Nanomaterials (2) and, for a critique of diamond synthesis as a starting point — which, contrary to legend, I’ve never advocated — see Nanomaterials (1).

It should also be noted that there isn’t any necessity for using diamond-like materials for anything but the most extreme applications. It’s just that diamond-like materials are especially good for many purposes, and will, in fact, be accessible — eventually.

The controversy about these materials has been a distraction. I wouldn’t have said so much about them if they hadn’t been so easy to analyze computationally, using standard molecular mechanics models, and hadn’t later come under attack by critics of my work. As often happens, the fiercest controversies have centered on questions that aren’t important.


See also:


{ 8 comments… read them below or add one }

Guy December 20, 2009 at 7:26 pm UTC

Glad I could provide food for thought.
However, recent work in functionalized adamantane chemistry does provide at least the potentiality for rigid, carbon rich structures from organic reaction environments. I’m not talking about a finished assembler system here mind you (neither was I in my last comment) but an intermediary device which produces perhaps only a single unique but previously unproducible structure such as diamond nanowires or custom length polymers.
Yes, yes, I know “speculation” but speculation is all I’m being paid to do.

Eric Drexler December 20, 2009 at 8:41 pm UTC

Along those lines, it’s notable that some proteins are kinetically stable in dry organic solvents, and that enzymes are now used under those conditions in industrial processes. Many of the supposed limitations of biomolecules are more myth than reality.

Chris Phoenix December 21, 2009 at 12:28 am UTC

Is it possible that no one will bother to build general-purpose manufacturing systems until it actually does become easy? I’ve posted a few speculations on my blog.

Valkyrie Ice December 29, 2009 at 6:23 am UTC

@Chris: Actually with the work being done on 3d printers, electronic printing and even organ printing, I think that we are moving towards a macroscale general assembler paradigm. I think it is likely that macroscale objects will be primarily made via printing by the end of this decade, with the continuing development of nanoscale manufacturing primarily in the construction of VLSI chips, as well as increasingly sophisticated medical applications which will spur more work on the integration of biosynthesis and electronics manufacture.

By the time we can routinely use nanoscale construction techniques to create a general purpose assembler, I suspect we will already have a robust macroscale version of the desktop manufacturing plant in many peoples homes, with a thriving open source community of devices it can make. From there it should be just a matter of a few years until those macroscale devices will be replaced by nanoscale versions, and just a few years after that before we begin seeing sophisticated nanoscale “modules” which can be used to design almost any product.

Eric Drexler December 30, 2009 at 4:34 am UTC

@ Valkyrie Ice — I’d be surprised to see printing methods become dominant, simply because specialized, high-throughput manufacturing technologies are so effective. These (which are mostly about non-nano manufacturing) show what I mean:

High-Throughput Nanomanufacturing: Small Parts (with videos)

High-Throughput Nanomanufacturing: Assembly (with videos)

An entertaining sample, with music:

Valkyrie Ice January 1, 2010 at 8:24 am UTC

I would agree with you on the speed and efficiency of dedicated machines, but I think the trend is going to be shifting away from high volume and relatively identical manufactured items to a need for smaller volumes with immensely more flexibility.

Look at Smartphones. They just released new models six months ago, then Droid a month ago, and now Google is releasing yet another new generation. Same with Computers, with Manufactures outsourcing actual construction to companies that make items for a dozen competing brands. We are developing new models with more features in less time, and fewer sales are happening per variant while the number of choices are growing constantly. Generation times are becoming so fast that a million units may not even have time to move off the shelf before the next better thing is out. That seems to indicate that generation times will force the production of fewer units, with a drive to minimize retooling times for assembly lines, which is likely to lead to a need for robust general manufacturing units that can switch production lines with minimal time.

Also as printing of electronics allows moving away from the hard fiberglass circuit board to materials that can be run roll to roll, the time between designs will likely also begin changing even more drastically than it currently is. We may soon arrive at a point where new generations might take weeks or days, demanding manufacturing capabilities which can keep up with the speed. Only general purpose, extremely versatile manufacturing systems would be able to do so.

Speed is becoming far too important to the manufacturing times to allow a continuation of single purpose devices, even though they may be far faster per step in a manufacturing process. They will simply take to long to design for a era in which the technology can change before the assembly line could be built.

The big electronics companies seem well on their way to becoming design studios, and the only way the manufacturers will be able to keep up is by embracing general manufacturing devices, and speed, over specialized, and efficiency.

Eric Drexler January 2, 2010 at 7:13 am UTC

@ Valkyrie Ice — I think we’re looking in the same direction, but with different emphasis on what we see. Downstream, high-throughput specialized machines will become very small and cheap (and atomically precise), and the main advantages of high specialization then, as now, will be in producing parts that can serve as standard components across wide range of products (screws, sockets, microprocessors…). The most robust advantages of generality are in putting standard components together in new, perhaps unique ways. A natural application of unspecialized machines that make (not just assemble) components is production of potentially-unique structures at the scale of the whole system, giving it form — for example, casings and frameworks. These can often be relatively low in precision, materials properties, and structural complexity, yet play an essential role in making a new product that, internally, might be a unique combination of very high technology parts. I think that all this is compatible with what you’ve described.

Valkyrie Ice January 3, 2010 at 2:25 am UTC

Ah, I see. indeed. I do think that modular design is going to become a very common part of production, especially in an open source environment, in which high knowledge designers make basic modules and test for safety, building an ever growing database of molecular designs for the less technically knowledgeable to use in novel designs, and I can see dedicated production units for those modules mixed with a general purpose assembler to combine those components in any configuration.

It really depends on how sophisticated such projects as RepRap become. That many companies are working on electronics printing is also going to become quite a factor. Since printed electronics can be printed onto a wide set of substrates, it’s going to begin spreading electronics into areas which have never previously been permeable to electronics designers. I would not be surprised to see a form of video wallpaper in less than two years.

Should be interesting to watch how the market develops.

{ 1 trackback }

Leave a Comment

Previous post:

Next post: