FAI FAQ draft: What is nanotechnology?

by lukeprog1 min read22nd Nov 201123 comments

3

Nanotechnology
Personal Blog

I invite your feedback on this snippet from the forthcoming Friendly AI FAQ. This one is an answer to the question "What is nanotechnology?" For references, see here.

_____

 

Nanotechnology is the study of materials and devices built at the scale of 1-100 nanometers (“nano-” means “one billionth of”). A hydrogen atom is about 0.24nm across, so we’re talking about materials and devices built atom by atom.

One famous piece of nanotechnology is the carbon nanotube. A carbon nanotube is a one-atom-thick sheet of graphite that is rolled into a seamless tube. Because of their physical properties, carbon nanotubes usually allow ballistic conduction, meaning that electrons can flow through the tube without collisions (Lin & Shung 1995), which allows the carbon nanotubes to conduct electricity without heat dissipation (Chico et al. 1996)! Carbon nanotubes are also much stronger than diamond or steel (Popov et al. 2002). Easton Bell Sports uses carbon nanotubes to build tougher bicycles, doctors use carbon nanotubes as scaffolding for bone growth in tissue engineering applications (Zanello et al. 2006), and one company uses carbon nanotubes to produce a special kind of high-conductance heater.

New nanomaterials are being developed every year, and may see applications in nearly every field of technology (Allhoff et al. 2010). Nanotechnology has already given us stain-free pants, larger-capacity hard drives, stronger cement, longer-lasting tennis balls, the world’s first sale of a quantum computer, a new method for fighting cancer, and much more.

An even more radical technology was described in Eric Drexler’s (1987) Engines of Creation. As Allhoff et al. (2010, p. 7) explain, Drexler predicted

a new form of technology based on molecular “assemblers,” which would be able to “place atoms in almost any reasonable arrangement” and thereby allow the formation of “almost anything the laws of nature allow.” This may sound like a fanciful and fantastical idea but, as Drexler points out, this is something that nature already does, unaided by human design, with the biologically based machines inside our own bodies (and those of any biological species).

Tiny molecular machines called “nanobots” would be a particularly revolutionary invention. For example in nanomedicine they would allow us to intelligently access cancer cells and blood cells.

It may also be possible to build self-replicating nanobots. These nanobots would use materials in their environment to manufacture copies of themselves. This would be an explosive technology, as Drexler (1987, p. 58) explains:

[The] first replicator assembles a copy in one thousand seconds, the two replicators then build two more in the next thousand seconds, the four build another four, and the eight build another eight. At the end of ten hours, there are not thirty-six new replicators, but over 68 billion. In less than a day, they would weigh a ton; in less than two days, they would outweigh the Earth... if the bottle of chemicals hadn't run dry long before.

But the massive production of nanobots does not require that they be self-replicating. They could also be produced by nanofactories (Phoenix 2005).

Access to (or invention of) advanced molecular manufacturing is one thing that could make a machine superintelligence incredibly powerful.

3

23 comments, sorted by Highlighting new comments since Today at 11:55 PM
New Comment

Because of their physical properties, carbon nanotubes usually allow ballistic conduction, meaning that electrons can flow through the tube without collisions (Lin & Shung 1995), which allows the carbon nanotubes to conduct electricity without heat dissipation (Chico et al. 1996)!

This sounds a bit sensationalist. The caveat is that this only happens on the scale of 1 micrometer at best.

Also with reference to that quote, I think you (Luke) should use a period rather than an exclamation mark. The exclamation mark makes it sound a bit unprofessional, and also looks a bit dumb because of the parentheses.

Overall, I think it looks pretty good.

a new form of technology based on molecular “assemblers,” which would be able to “place atoms in almost any reasonable arrangement” and thereby allow the formation of “almost anything the laws of nature allow.”

This really just sounds like we're talking about the ability to create extra strong metal, or scaled up versions of the "stain-free pants" you mentioned, which doesn't really cover why MNT would be so powerful. For someone who has never heard of this subject, it will take more information before they realize why this is more important than, say, a new method to create higher resolution graphics on computer screens. Explain why MNT is worth discussing in a Singularity FAQ, instead of any of the other classes of technology that look feasible.

This should probably at least mention MNT computers and MNT weaponry. Nanofactories/assemblers could make a lot of stuff other than just nanobots. The emphasis on nanomaterials seems misplaced in that they have little to do with FAI. Maybe mention MNT critics and make some attempt at defusing them.

For example in nanomedicine they would allow us to intelligently access cancer cells and blood cells.

Needs comma.

This FAQ suffers a compression fallacy between chemistry (the products you've named that actually exist) and nanoscopic robots (which don't, and we have little to no idea how to get there from here, and the named products of chemistry don't actually help). Crediting the first as examples of the second is the sort of thing that gets chemists I know punching walls. I suggest taking care not to conflate chemistry with nanoscopic robotics.

Isn't an H atom more like 0.1nm in diameter? Of course it's fuzzy.

I agree with steven0461's criticisms. Drexler outlines a computer design giving a lower bound of 10^16 instructions/second/watt.

Should there be a ref to http://e-drexler.com/d/07/00/1204TechnologyRoadmap.html ?

Quibbling about words: "atom by atom" seems to have caused some confusion with some people (taking it literally as defining how you build things when the important criterion is atomic precision). Also "nanobots" was coined in a ST:TNG episode, IIRC, and I'm not sure if people in the field use it.

You're thinking of "nanites", I'm pretty sure.

Oh, you're right, thanks.

Apparently .24nm is twice the Van der Waals radius and .1nm is twice the Bohr radius. I'm not sure which one has a better case for being called the "true radius".

Ah -- .1nm is also the C-H or C-C bond length, which comes to mind more naturally to me thinking about the scale of an organic molecule -- enough to make me wonder where the 0.24 was coming from. E.g. a (much bigger) sulfur atom can have bonds that long.

Drexler outlines a computer design giving a lower bound of 10^16 instructions/second/watt.

I've seen this before but now I can't find it. Do you have a link?

I think it's in Chapter 12 of Drexler's Nanosystems, which may be worth citing regardless. ETA: or to read about it online, Chapter 11 of his MIT dissertation linked at the top.

Yes, that's where I got the figure (the printed book). The opening chapter lists a bunch of other figures of merit for other applications (strength of materials, power density, etc.)

I'll note that something that took me a long time to understand about nanotech in my relative youth was that the difference in scale between nanotechnological structures and cellular structures mattered, which meant that nanotech was meaningfully different from biotech. (Roughly speaking, I think I was lumping all "very small things" into the same bucket.) Assuming that wasn't a complete idiosyncracy of mine, you might want to emphasize explicitly the important differences between custom-built self-replicating nanobots & custom-built self-replicating bacteria.

This is important, as many nanotechnology advocates (starting from Drexler himself) point to life as proof by example that what they're trying for is possible. Though this is a less egregious conflation than seguing from chemistry (new materials) into nanoscopic robotics.

(“nano-” means “one billionth of”)

http://lesswrong.com/lw/fk/survey_results/

IQs (warning: self-reported numbers for notoriously hard-to-measure statistic) ranged from 120 to 180. The mean was 145.88, median was 141.50, and SD was 14.02. Quartiles were 155.

(“nano-” means “one billionth of”, so a nanometer is one billionth the size of a meter).

allow us to intelligently access cancer cells and blood cells.

I don't know what this means.

This would be an explosive technology

I can't think of a better word than "explosive", but we should try.

IQs [...] ranged from 120 to 180.

I think this FAQ is intended for a general audience.

[This comment is no longer endorsed by its author]Reply

That's why I think it should be said that a nanometer is a billionth of a meter and not just that nano means "one billionth of." I thought the original was overestimating the audience. Apparently I communicated poorly as I was interpreted to mean the opposite.

Ah, gotcha.

But the massive production of nanobots

This sounds weird to me. "The massive production" weakly implies complete confidence that a specific event of massive production will actually occur. "But massive production of nanobots does not require that they be self-replicating," would be better, or "But massive production of nanobots would not require that they be self-replicating," as it is hypothetical.

"Massive production" is distracting because of the similarity to "mass production," so maybe say "extensive production" instead.

"The production of large numbers of"?

It is worth mentioning the possibility of macro-scale self-replicators here, or are they an entirely different topic?

[-][anonymous]9y 0

Where is all soft nanotechnology? There is sometimes a distinction between soft and hard nanotechnology, Drexler reprecenting hard.

This is not really my subject but from what I've read there seems to be a lot of nano-tech research (aside from nanotubes) that is focused on self-assembly - not as in "nanobot-self-replication", but more like protein folding, what would be classified as soft.

Edit: I just realized that the kind of of nanotechnology that is really relevant for the FAQ is the one Drexler is proposing, but it seems to me as the description is analogous to trying to explain the singularity but only describing Kurzweil's definition.