From RationalWiki
(Redirected from Nanobot)
Jump to navigation Jump to search
Comparison, in nanometers, of various nanomaterials and some items that are visible to the naked eye.
We need the best
Icon Tech Portal.svg
Programming for Dummies
It matters
Icon chemistry.png
Action and reaction
Spooky scary chemicals
Er, who's got the pox?

Nanotechnology (often referred to as "nanotech", or sometimes "nanoscience") is the art of working on materials at the nanometer (one billionth of a meter) scale, hence the name. For perspective, a human hair is about 80,000 to 100,000 nanometers wide.[1]

There is no single unified thing that can be called "nanotechnology". Rather, many different technologies and industries are reaching the point where nanoscale operations are of interest. In previous years, we might have called this "chemistry" or "materials science", though a lot of this is now being called straight-up "nanotechnology".

There are two sorts of nanotechnology: reality, and science woo.

In practical terms (i.e., funding it), it's chemistry with "nano-" on the front because that gets funding from people who think they're paying for magical molecular robots. Emphasis on "magical". This is improving a bit, as chemistry funders realise the non-magical version is worth it in itself. Real nanotech is great stuff!

Unfortunately, advocates of woo nanotechnology will regularly push a future of magical molecular robots, then point at chemistry as evidence for the magical molecular robots.


The idea was first postulated by Richard Feynman in his 1959 essay There's Plenty Of Room At The Bottom.[2] The talk was not widely noticed and did not result in any actual follow-up research; it was rediscovered and used to promote the field exploiting Feynman's fame.

The scanning tunnelling microscope (STM) was invented in 1981, which enabled microscopic imaging of individual atoms, and a STM was first used to manipulate individual atoms in 1990. In the 2000s and 2010s, advances in atomic force microscopy (AFM) led to the first pictures of individual orbitals. None of this progress yet means that nanorobots could be mass-produced. Nanomachines have been actually made with the methods of conventional organic synthesis, but this term is not as sexy as "nanotech".

From the late 1980s on, it was popularised by K. Eric Drexler, in his 1986 book Engines of Creation: The Coming Era of Nanotechnology and the peer-reviewed work of "theoretical applied science" exploratory engineeringWikipedia and computer science that Drexler used as a Ph.D. thesis. (Drexler's Ph.D. is much-touted as being from MIT, but was not awarded by an official science or engineering department — he got it from the Media Lab, an interdisciplinary department typically focused on art and biomimetics research).[3]

Nanotechnology fanboys — as opposed to the people who actually work with the stuff — have a habit of downplaying[citation needed] Feynman's origination of the idea and playing up Drexler, possibly because the latter is far more indulgent of their fantasies.

Dr. Drexler, most upset that the National Nanotechnology Initiative insisted on only funding things that are profitable,[4] has taken to using his new favored terms, "molecular nanotechnology" or "atomically precise manufacturing".[5]

In science woo[edit]

No, you don’t get it. You are still in a pretend world where atoms go where you want because your computer program directs them to go there.
—Richard Smalley to Drexler

The popular conception of nanotechnology is Eric Drexler's concept of nanobots, like industrial robots scaled down a million times. This is entirely made of bollocks and would violate physics, chemistry, and thermodynamics.[6][7] However, not many people realise why it's bollocks, as it isn't really due to us limiting what our technology might one day do, it's purely down to the physics of how the world works.

The rules of mechanics don't scale evenly with size — consider surface to volume ratios, which are why the energy consumed by small rodents per unit of mass is different from that for elephants, and why the aerodynamics of a Boeing 747 are different from a small 747-shaped toy. At the molecular level, the world is considerably different from the macro scale. A solvent like water would feel more like treacle, and anything floating in it would be at the mercy of Brownian motion that would make even the strongest rip currents look quite weak.

Imagine what manufacturing would look like if your machines' gears rotated randomly, the size of the gears' teeth fluctuated randomly, everything (gears, grippers, working materials) was magnetized, and the whole factory burst into flames because of ionizing radiation. After realizing this, Drexler proposed "machine-phase chemistry",[8] an entirely hypothetical state of matter that avoids these problems through extreme stiffness and superlubricity.[9] Quantum chemistry simulations have tested machine-phase matter, and confirmed it can probably manufacture more machine-phase matter.[10][11][note 1] At least the US patent office thinks so![note 2]

The chicken-and-egg problem: we don't know of any way to transform solution-phase matter into machine-phase. Even though we can simulate this exotic state of matter, it is entirely theoretical and we don't know for certain how it will actually behave in an experiment. Maybe it develops a mind of its own and instantly self-destructs! We just can't study it in anything beyond approximate physics simulations.

Once in machine-phase, a universal constructor would self-replicate[note 3][note 4] until there were enough nano-factories for every person on Earth. This would supposedly lower the cost of manufacturing significantly. However, we saw the same prediction come about with 3D printing. People actually tried this concept with the RepRap projectWikipedia — and it promptly went out of business. Until proven otherwise, we cannot assume a nano-factory will turn out differently.

That Drexler has pushed a design which is impossible to build has not deterred nanotech fans, who get as upset as they usually do when an expert in a field they're talking about points out they're simply wrong, apparently from an emotional sunk cost fallacy.

Transhumanists routinely casually throw biology out the window, but with nanobot advocacy, they manage the same contempt for chemistry and physics. (Michael Anissimov, then of SIAI, advocated Digital Rights Management as the way to control dangerous science woo nanotechnology,[12] thus also casually throwing mathematics out the window.) In cryonics, "but, nanobots!" is the standard answer to any objection. Some cryonics advocates are finally pointing out that this is not quite rational.[13]

Advocates of the woo version of nanotech often tag life "soft nanotech", as if that makes "hard nanotech" (woo) merely unrealised rather than impossible. This is in reference to the fact that such "hard" nano-tech seems to imply solid, robot-like structures (made of diamond or cerium dioxide[14]) while "soft" is closer to more malleable molecular structures. It's a bit like calling speechWikipedia "soft telepathy" to imply that "hard telepathy" is a difference of degree, rather than kind. Drexler has recently advocated for ionic compounds in the "hard nanotech" category, citing that biological organisms already build such crystals.[15] Pyrite is the best material according to his new performance metric, but easily transforms into toxic sulfur dioxide and hydrogen sulfide. He instead advocates for cerium dioxide, although long-term exposure to Ce3+ ions is cytotoxic.[16]

Additionally, going back to the problem of scale for a moment, consider the following points:

  • Each individual nanobot would need it's own tiny power source, but it's pretty hard to imagine, say, a Duracell battery composed of only twelve or so atoms functioning properly.
  • How would the nanomachines perceive their surroundings? Even assuming that a camera lens composed of only a handful of atoms even interacts with photons the same way a macroscopic one would, it would simply be too small to have the domed shape needed to focus light peroperly.

In fiction[edit]

Nanomachines, son!
—Senator Steven Armstrong explaining his absurd resilience to Raiden, Metal Gear Rising Revengeance

Drexler's vision of nanobots is a standard science fiction trope for magic. Applications include superhuman healing and body repair, superhuman powers, shapeshifting machines, constructing large structures at a miraculous pace,[17] self-repairing materials, and instantly taking over and reprogramming any computer system[18] — or even a person.[19] Alternatively, a swarm of nanomachines can be a weapon, eating anything in its path like locusts.[20]

One common version is self-replicating nanomachines, which convert any available matter into more nanomachines to replenish and bolster their numbers. If left unchecked, these machines convert everything — the so-called "grey goo" doomsday scenario.[21][note 5] The idea of creating self-replicating machines, which was an early goal of nanotechnology, has given way to the safer and simpler approach of building larger (relatively speaking) factories that can mass-produce nanomachines, making this scenario less likely to occur in reality.[23][24][25] That said, the basic concept of self-replicating machines is solid in itself and some research has been done on creating them on a larger scale, and cellular automata like von Neumann's universal constructor can be argued to act as an abstract simulation of a Drexler-style replicating nanobot.[26]

Writers like these little things because they don't have to show and explain how they work — all you see are the results.

A slightly more believable description of nanotechnology can be found in Neal Stephenson's book The Diamond Age.Wikipedia In this, not only is there a detailed description of how the machines would be built, there is even a discussion of associated problems, such as heat dissipation.

In reality[edit]

Reality is not as fantastically exciting as the fiction — but it's still pretty cool.

A chemical diagram of a carbon nanotube.
A chemical diagram of C60 (Buckminister Fullerene).

Current nanotechnology focuses on creating much simpler structures, such as nanoparticles that can deliver drugs to specific cells, and materials with advantageous properties. Better understanding of materials at the nanoscale is explaining many apparently odd things. Nanoparticles (colloids) of gold are red, rather than yellow as the bulk metal is, and this effect is responsible for the colours of medieval stained glass. Some of the odd properties of soot are because it contains buckyballs (a.k.a., Buckminister Fullerene) and nanotubes. Scientists working on these nano-scaled advances tend to refer to it as either "chemistry" or "materials science", as "nanotechnology" is not a functioning research field in its own right and is just an umbrella term for a variety of research ventures that span multiple disciplines (and a handy buzzword to get funding from people who think they're buying magical robots).

Molecular recognition — the ability for catalysts to actively recognise and orientate their substrates — is an emerging area of research and could become the norm for chemical synthesis in a few decades. However, this still relies on normal chemical synthetic routes with one catalyst tuned to one reaction (or group of reactions).

Molecular nanotechnology, which aims to engineer mechanical systems at the molecular level, is barely in its infancy. The dream of assembling molecules "atom by atom" may not even be possible in the sense claimed by those writing about nanobots. So far, scientists are pretty good at arranging atoms in patterns with a scanning tunneling microscope (STM), as was famously done by IBM to demonstrate STM technology, and making "quantum corrals" that demonstrate the wave-like properties of electrons. If nanomachines are built, they will work much more like the currently-known nanomachines — antibodies and proteins and so forth, restricted to catalysing only one family of reactions — and more complicated nanomachines will be closer to the size of biological cells.

Mechanical nanocomputers are theoretically possible, and research is steadily getting there. So far, there's a 300nm electromechanical reed relay gate[27][28] and an inverter that runs at 500kHz.[29] The application is for environments that would trash electronics, e.g. high temperatures. For comparison, current computers' electronic transistors are on the order of 7nm (as of 2017)[30][31][32] and mainstream consumer computer chips run at between 1GHz and 4GHz or so.

That's not to say that this stuff isn't insanely cool. For instance, sending in a specially-designed killer molecule to cure cancer.[33] Holy crap!

There are examples of self-replicating machines that dig into the ground and vacuum up the atmosphere. They sense the available resources of the surrounding territory and assemble useful products, even erecting mini solar panels to assist the process as well as producing the next generation of their initial seeding mechanism, all fully automated and largely unattended. They are called things like "tomato plants".

Nanoparticles of titanium dioxide are used to whiten some food products like jawbreakers/gobstoppers.

Nano-sized particles (<100 nm) have been used in consumer products, including clothing, electronics, cosmetics, and food, for at least two decades. The safety of nanoparticles in food is not well understood, and is still being investigated, though modest amounts are believed to be safe. In the US, nanoparticles in food are currently regulated the same as any chemical additive. Titanium dioxide, silicon dioxide, and zinc oxide are currently the most common nanoparticles added to food.[34]


Tiny self-replicating nano-machines have already existed for millennia. They are called viruses. As you can plainly observe, even though viruses are rapidly evolving and recursively self-improving, they haven't ended all life on Earth. Spreading the myth that self-replicating nano-machines are a future new super-technological threat only causes unjustified fear, uncertainty, and doubt.

Viruses have killed millions of humans, but thanks to advances in medical science (especially vaccines), we can fight back. Humanity has wiped out smallpox.


Drexlerian nanotechnology as an idea doesn't have any problems per se, particularly as Drexlerian nanotechnology doesn't exist. However, there are problems with some of the public depictions or expectations of nanotechnology, especially if Drexler-style nanotech is portrayed as "just around the corner" instead of part of the abstract sci-fi realm along with things like anti-gravity.

The main practical problem with nanotechnology is the phrase being overused as funding hype for what anyone else would call materials science or synthetic biology.


Nanoparticles are small enough to mess with biological processes.

The mechanism of carcinogenicity for asbestos is frustrated phagocytosis, i.e., the inability of a phagocyte to engulf its target.[35] This mechanism likely also applies to other similarly-shaped, biologically-inert substances such as nanotubes and nanowires that meet the same length threshold for pleural retention of 5 µm, as asbestos does.[36] In 2017, The International Agency for Research on Cancer concluded that some types of multiwalled carbon nanotubes (specifically, MWCNT-7) are carcinogenic in experimental animal studies.[37]

However, functionalized carbon nanotubes have shown biodegradability.[38] This means nanomechanical elements littered with functional groups may be fully biodegradable, although known technology cannot create and test such components. This necessitates designing a nanobot so its hard shell eventually self-destructs, exposing the fragile nanomachinery.[39] An extremely dense form of rod logic would permit RISC computers the size of 100 nm, so nanobots could be small enough to remove the risk of asbestosis entirely.[note 6] However, they could cause harm through other mechanisms. One example is nanobots walking on cellular plasma membranes, then accidentally sending mechanical signals through the extracellular matrix.[40] The human body uses mechanical signals for numerous purposes, including apoptosis (cell death).[41] This may be useful for cancer cells, but definitely not healthy cells!

Cadmium and some cadmium compounds are carcinogenic to humans.[42] Quantum dots contain cadmium and there has been work to reduce the potential toxicity of quantum dots.[43]

Eva Oberdörster reported that fullerenes (C60) in colloidal form caused oxidative stress in juvenile fish at a concentration of 0.5 ppm.[44]

More generally, nanoparticles that occur in pollution, known as "fine particles" (<2.5 µm wide), are known to cause both respiratory and cardiovascular health problems.[45] There is also some evidence that inhalation of fine particles cause other diseases, including diabetes, obesity and dementia.[45] A study showed that humans who inhaled relatively-inert gold nanoparticles had the particles in their bloodstream after 15 minutes and the particles remained in the body for as long as 3 months.[45] This study demonstrated the mechanistic plausibility of inhaled non-inert types of nanoparticles causing non-cardiovascular/non-respiratory diseases.

Conservative estimation[edit]

Computational chemistry simulations can only simulate a finite amount of atoms, and faster simulations sacrifice some amount of accuracy for speed. Conservative estimation: to compensate for our lack of knowledge, assume the worst will happen. A single cosmic ray hits a component of a nano-factory — the component breaks beyond repair.Wikipedia In fact, an entire trail of components and anything touching them also breaks beyond repair. This would seem to make any system with such fragility unworkable. The math gives a surprising answer: a 400-nm cubic object would survive ionizing radiation for 100 years. For comparison, the best quantum qubit lasts two milliseconds[46] and still does something useful. It doesn't take much imagination to apply redundancy,Wikipedia a standard engineering practice, to the design of a nano-factory.

Yet, without taking a minute to check the math, it's easy to claim this is a fatal flaw,[note 7] and say the entire thing defies physics. Some people (e.g. Richard Smalley) top it off by ridiculing the author. This discourages the audience from checking the math; the ridicule amplifies a pre-conceived notion that it will be heavily flawed. This is not scientific, because the criticisms aren't based on technical aspects of the design. The same tactic effectively scares away even intelligent people from considering nuclear energy. People assume that the nuclear power plant will blow up like Chernobyl, overlooking the mathematically sound fact that nuclear energy saves lives by displacing CO2 emissions. This also suppresses genuine criticism such as nuclear energy's inability to deploy rapidly like solar power.

The lack of technical criticism isn't a green flag that the design obeys physics, either.[note 8] Each specific risk needs to be considered and quantified in the engineering constraints. This is exactly what Drexler did in Nanosystems, and it led to some counterintuitive insights.[note 9] The concept of nuclear power arose in a very similar manner. It was soon touted (c. 1950)[citation needed] as something that would bring universal cheap energy to everyone, before the government made it absurdly expensive. Drexler's nano-factory could have cool uses (like femtotechnology nuclear technology), but it may not be as fantabulous as some people expect. It also won't cover us in grey goo, just like nuclear technology hasn't (yet) covered the world in ash.

See also[edit]

External links[edit]


  1. Under strict human control. There is a difference between autonomy and self-replication… at least we hope.
  2. The office has also patented perpetual motion machines, although Zyvex's research has decent scientific rigor and anyone can reproduce it with a personal computer. It also obeys the laws of thermodynamics.
  3. Not self-replication like molecular assemblers in Engines of Creation. This is a chiefly different kind, the one already happening with modern computers. A computer designs and performs the manufacturing of CPUs, which then become computers. The nano-factory would technically self-replicate in a similar manner, with significant assistance from human beings.
  4. There is also a major difference between a centimeters-sized nano-factory replicating (potentially plausible) and a 100-nm-sized nanobot self-replicating (even Drexler says that's impossible).
  5. Even if it did happen, grey goo (also called eco-phagy) would not cause human extinction: "Ecophagy that proceeds slowly enough to add ~4°C to global warming (near the current threshold for immediate climatological detection) will require ~20 months to run to completion; faster ecophagic devices run hotter, allowing quicker detection by policing authorities. All ecophagic scenarios examined appear to permit early detection by vigilant monitoring, thus enabling rapid deployment of effective defensive instrumentalities."[22] Their thermodynamic requirements mean they'd also be quite vulnerable and easy to thwart.
  6. Nanobots larger than 2 microns will clot capillaries; Freitas's Nanomedicine seems to only care about this size constraint.
  7. Or, like transhumanists, claim the opposite while exercising an equal lack of scientific effort.
  8. And is not proof that it can be built.
  9. One great example that is often overlooked: nanomechanical computers would be more robust than nanoelectronic computers. We might even have a way to build one with modern technology: superlubricant carbon nanotubes. This conclusion came from assuming our computational models wouldn't be able to model electronics as well as nanomachinery based on classical physics. Therefore, we would not be able to design good nanoelectronic circuits. Since 1992, the state of modern computational chemistry algorithms seems to reflect this assumption. DFT electronic structure calculations are O(n3-4) while molecular dynamics is O(n1-2).


  1. "Size of the Nanoscale". National Nanotechnology Initiative.
  2. Richard P. Feynman (December 1959) "Plenty of Room at the Bottom"
  3. Original thesis PDF.
  4. Let the Nanotech Wars Begin! (KurzweilAI, 14 December 2003)
  5. Exploratory Engineering and Radical Abundance: an Exclusive Interview with Dr. Eric Drexler by Paul Raven & Eric Drexler (November 30, 2011) h+ Magazine (archived from January 23, 2012).
  6. Rupturing The Nanotech Rapture: Biological nanobots could repair and improve the human body, but they'll be more bio than bot Richard A.L. Jones, IEEE Spectrum, June 2008
  7. Nano-nonsense: 25 years of charlatanry (Scott Locklin, Locklin on science blog, August 25th 2010)
  8. Mechadense's Wiki: Machine phase
  9. Superlubricity in centimetres-long double-walled carbon nanotubes under ambient conditions
  10. Simple tool for positional diamond mechanosynthesis, and its method of manufacture
  11. Positional diamondoid mechanosynthesis
  13. Cryonicists' sunk costs reasoning about Drexler's "nanotech." (Mark Plus, The Life of Man Qua Man on Earth blog, 2010-08-31)
  14. Metamodern: Nanomachines, Nanomaterials, and K_lm
  15. Why diamond synthesis is a bad objective
  16. Cytotoxicity and Genotoxicity of Ceria Nanoparticles on Different Cell Lines in Vitro
  17. Total Annihilation: "It would have been too complex and time consuming to have little guys with hammers and scaffolds every time something was built in the game. It also wasn't futuristic enough. We needed something like magic, but with a thin veneer of science around it. Nanotechnology to the rescue!"
  18. T-XWikipedia
  20. The Day the Earth Stood Still (2008 film),Wikipedia G.I. Joe: The Rise of Cobra,Wikipedia etc.
  21. For example, the Replicators in Stargate.
  22. Some Limits to Global Ecophagy by Biovorous Nanoreplicators, with Public Policy Recommendations
  25. Nanotechnology pioneer slays 'grey goo' myths
  26. Kinematic Self-Replicating Machines
  27. Babbage nanomachine promises low-energy computing (New Scientist #2573, 25 March 2010)
  28. Diego N. Guerra, Adi R. Bulsara, William L. Ditto, Sudeshna Sinha, K. Murali and P. Mohanty, A Noise-Assisted Reprogrammable Nanomechanical Logic Gate. Nano Letters, 10 March 2010, DOI: 10.1021/nl9034175
  30. IBM Research builds functional 7nm processor
  31. IBM Discloses Working Version of a Much Higher-Capacity Chip -
  32. By 2020, there should be 5nm transistors, and in 2008, a team of UK researchers made a transistor 1 atom thick and 10 atoms wide out of graphene.
  33. This Is the Future of the Fight Against Cancer (Janet Fang, Nature, 21 March 2010 doi:10.1038/news.2010.138)
  34. Nanoparticles in foods raise safety questions: These microadditives enhance color, flavor and freshness. But what do they do in the body? by Susan Gaidos (1:28pm, October 16, 2015) Science News.
  35. IARC Monographs on the Evaluation of Carcinogenic Risks to Humans: Asbestos (Chrysotile, Amosite, Crocidolite, Tremolite, Actinolite and Anthophyllite)
  36. Use of back-scatter electron signals to visualise cell/nanowires interactions in vitro and in vivo; frustrated phagocytosis of long fibres in macrophages and compartmentalisation in mesothelial cells in vivo by Anja Schinwald and Ken Donaldson (2012) Particle and Fibre Toxicology 9:34.
  37. Some Nanomaterials and Some Fibres] (2017) IARC Monographs, Volume 111.
  38. Functionlized carbon nanotubes: biomedical applications
  39. Nanomedicine, Volume IIA: Biocompatibility (summary of potential biocompatibility issues when designing medical nanobots or nanoparticles)
  40. Nanomedicine, Volume 1: Basic Capabilities - Cytoambulation
  41. Mechanical Regulation of Apoptosis in the Cardiovascular System
  42. Agents Classified by the IARC Monographs, Volumes 1–118
  43. Synthetic Developments of Nontoxic Quantum Dots by A. Das & P. T. Snee (2016) Chemphyschem. 17(5):598-617. doi: 10.1002/cphc.201500837.
  44. Manufactured Nanomaterials (Fullerenes, C60) Induce Oxidative Stress in the Brain of Juvenile Largemouth Bass by Eva Oberdörster (2004) Environ. Health Perspect. 112(10): 1058–1062. doi:10.1289/ehp.7021
  45. 45.0 45.1 45.2 The list of diseases linked to air pollution is growing: As governments decide what to do about air quality, studies connect an array of health problems to dirty air by Laura Beil (7:00am, September 19, 2017) Science News.
  46. For the longest time: Quantum computing engineers set new standard in silicon chip performance