Lessons from History

An article in the December 2003 Smithsonian contains a cautionary tale for U.S. nanotechnology efforts. The article, "Taking Wing: A Century of Flight", tells how America lost the lead in aviation in the decades following the Wright brothers' flight.

William Pickering, an eminent Harvard astronomer, is quoted as saying in 1908: "It is doubtful if aeroplanes will ever cross the ocean. The public has greatly overestimated the possibilities of the aeroplane, imagining that in another generation they will be able to fly over to London in a day. This is manifestly impossible."

Nineteen years later, Charles Lindbergh made the first solo transatlantic flight.

However, the article asserts that "Such disdain chilled U.S. investment in aviation. Between 1908 and 1913, the U.S. government spent only $435,000 on aviation -- less than Germany, France, Chile and even Bulgaria. European inventors and entrepreneurs were soon building better, faster and more stable planes than were the Wrights."

When jets were invented in England and Japan, "In America, military brass put jets on a back burner, convinced the war would be won with conventional airplanes, and lots of them. Diverting resources to work on the unproven jet, authorities insisted, would be a waste of time."

The world's first operational jet flew in the Luftwaffe. "But after the Allies swept through Germany at the end of the war, they recruited dozens of German jet and rocket scientists, including Wernher von Braun, and then took them to the United States in 'Operation Paperclip'."

One might imagine a current mainstream nanotechnologist saying, "It is doubtful if nanomachines will ever do chemistry. The public has greatly overestimated the possibilities of mechanosynthesis, imagining that in another generation nanofactories will be able to build large products in a day. This is manifestly impossible."

And one might imagine history recording that: "Such disdain chilled U.S. investment in molecular manufacturing. Between 1992 and 2004, the U.S. government spent only $2 million on molecular manufacturing -- less than Russia, China, and even Iran. Asian inventors and entrepreneurs were soon building stronger, more stable, and more complex nanomachines than were Smalley and Whitesides."

This revised story has an eerie ring to it: "In America, nanotechnology administrators put molecular manufacturing on a back burner, convinced the market would be won with conventional nanomaterials, and lots of them. Diverting resources to work on the unproven mechanosynthetic fabricator, authorities insisted, would be a waste of time."

Now let's take it just a few years into the future, say 2010 or 2015. In this scenario, the world's first operational nanofactory is used by Chinese forces in the Siberian campaign. "But after the Alliance swept through Russia at the end of the war, they recruited dozens of Russian and Chinese mechanosynthesis and nanomachine scientists and then took them to the United States in 'Operation Stapler'."

Perhaps the most unlikely thing about this tale is the possibility of the side without a nanofactory being victorious in war.

Chris Phoenix

Comments

[Mr. Farlops]

Well, unlike my dismissiveness in my comment to the last post here, maybe I should be worried about this.

It would probably be better for all of us, if the United States were to begin a "Nanhatten Project" now, in relative peace (Despite "police operations" in Iraq, Afghanistan and the War on Terrorism.), then in time of war with an evenly matched foe.

Lindbergh and his technicians and backers used the potential market of transcontinental package delivery as a way to pursuade investors to pay for his Atlantic flight.

Considering it briefly, I can't think of any obvious equivalent to convince investors to fund development of a nanofactory. I mean, many here know that MNT will revolutionize everything, but that's not so obvious to the skeptics. There is no immediately obvious market in which to apply the technology, that can be made plain to investors.

I think the reason for this is that MNT is so broad and powerful that it applies to everything, to all markets. This strength can be a weakness when seeking money from people and groups unfamiliar with MNT. Even MNT's medical applications, though the last word in medical technology, have rivals in convential, advanced technology.

But I am getting diverted here.

I am forced to agree that it's probably better in the long run for a democratic, stable and prosperous post-industrial country take the lead in nanofactory development. The United States, currently, looks like the best choice for this.

[Mike Treder, CRN]

Chris and I are obviously quite worried too. Otherwise we would not have founded CRN and spent thousands of hours working on it for no pay (Hint: you can always contribute to the cause).

You make a good point about MNT having such broad and deep application that it's hard to comprehend. Imagine trying to tell someone in 1750 what it would mean to have their whole house and their whole city using electrical power!

[Color Me Skeptical]

Molecular Assembly is the modern "AI".

Just as AI has been "just around the corner" forever, so too Drexlerian molecular assemblers are gonna be "just around the corner" for generations to come.

If you want molecular manufacture, tweak biological manufacture instead of going after the windmill.

[Flackbait]

Hmm... On one side, yay for nanotech. On the other side, why don't we just nuke the living hell out of them, bombard them with orbitally-dropped shaped kinetic slugs (i.e. drop space rocks or abalative-armor-plated equivalents) on them, etc etc? All that stuff we have without researching nanotech. And unless nanotech can come up with something that allows the troops of the Chinese army to survive the detonation of thermobaric bombs or tactical nukes over their troop formations (which is possible, but would be rather difficult to acchieve), these things can be easily trumped with technology we have now. Not that I'm saying "don't research nanotech, it's a waste of time." But don't build castles in the air, either-- you're just as dead if someone stabs you with a knife, and how long as that gem of technological acchievement been around?

[Shelby]

Flackbait:

Yes, a knife can still kill you. But better knives have not won any wars in a long time, because other stuff works better for killing people. The folks with the best military tech have a huge advantage in any conflict.

As for nukes, there are a few drawbacks. Not terribly pin-point (often important), environmental side-effects, unpopular with third parties, and likely to lead any nuclear-armed opponents to retaliate in kind. Nor are space-based rocks much use now -- they're incredibly expensive to raise to orbit, and we don't have the means to bring them from the moon or the asteroid belt.

None of which is to express a strong opinion on when nanotech will become military significant. It may be forty years, not ten (though I doubt it).

[kamil]

American military brass was correct when they gave priority to conventional aircraft. Jet engines at the time were a new technology that required very advanced alloys metallurgy, strategic metals, and maching that were not invented at the time or very expensive, the result was that no German “operational” jet aircraft had an engine lifetime of over 100 hours, and each engine was 10x cost of a conventional fighter aircraft engine a Me 262 used 2. Pilots had to be trained on this new type of aircraft. US had aircraft in development that were comparable to German types with greater but still insufficient operational qualities, what the US did not have was Hitler’s obsession with wonder weapons developed at great cost and put into service prematurely to the detriment of the war effort.

Kamil

[Chris Phoenix, CRN]

Kamil, good points about the American decision not to develop/deploy jets in WWII. Certainly Hitler was over-eager and over-ideological, and it cost him the war. But what about the earlier decades of lagging in airplane technology? My impression (feel free to correct me) is that early American airplanes were not very effective against the Germans, and we only had time to catch up because Britain held out through years of bombardment.

Chris

[Chris Phoenix, CRN]

Skeptical: AI was a very diffuse goal, but was not a technology. And it was based on trying to directly solve complex real-world problems in a new way without a full understanding of their complexity.

The goal of a tabletop nanofactory using mechanosynthesis is not comparable to AI. It is well-defined and can be decomposed into simple sub-problems. It does not deal with any "complex systems" in the mathematical sense. Complicated, sure--but predictable. Once built, it will have immediate application in building stuff that works by familiar engineering principles to solve familiar problems.

It's easy to take a skeptical attitude and make dismissive statements without foundation. But please explain, if you can, why AI is a better comparison for the nanofactory than aviation.

Chris

[Daniel Newby]

The argument here is historical science fiction. "If they had done this in the distant past, then these good things could have happened. Therefore you should accept my proposal that could make good things happen." That approach is fun, but predicting the past is ultimately pointless. One might as well ask "What if the Romans hadn't used lead for water and food vessels?" without bothering to check if the alternatives might have bankrupted them.

Chris said "The goal of a tabletop nanofactory using mechanosynthesis is not comparable to AI. It is well-defined and can be decomposed into simple sub-problems."

That's what the AI people said, before they got their collective ass handed to them. And they were right: all the little parts are straightforward exercises in computation. But big systems are more than the sum of their parts. Their failure modes, instability modes (in the control theory sense), operational logistics, and maintenance strategies are distinct from those of the component pieces. Accounting, politics, computer programming, digital chip design, drug discovery, you name it: perfect understanding of the components doesn't make you able to design complicated systems.

Chris said "Once built, it will have immediate application in building stuff that works by familiar engineering principles to solve familiar problems."

But it won't! The realm of the tiny is alien even to the best physical chemists and quantum physicists. They are blind men poking around with a blunt stick. Consider the problem of water wetting a glass surface. Nobody knows what the hell happens! Go ahead and ask a surface chemistry post doc. You'll get a lot of hand waving and general statements, and a bewildering pile of experimental data, but no answers you can take to the venture capitalists.

And silica and water are the two most common substances in chemistry! Unless you do your nanotech in a vacuum furnace, or have nano-bucket-brigades crawling around, you'll have to contend with a ubiquitous surface layer of water and/or hydroxyls. In fact, I would go so far as to say that if "hydroxyl" isn't in a person's working vocabulary, then they aren't qualified to be talking about molecular technology.

Chris said "It's easy to take a skeptical attitude and make dismissive statements without foundation."

Well, duh. Real scientists and engineers answer the objections, or say they don't know but it's plausible enough to try anyway. Historical analogy is not a meaningful response.

One more thing: Your entire argument is rooted in ignorance, from what I can tell. The relevant research you are wringing your hands about is being done at a breackneck pace. Surface chemistry (at this scale, it's *mostly* surface), active nanostructures (mechanical and computational), quantum mechanical modelling, supercomputers with fast memories, data warehousing, redundancy and failure mitigation, error detection and correction algorithms, nanomeasurement, and so forth. Billions of dollars of are being poured into the research. Just down the hall from my office are some projects working on making nanostructured materials, and the instrument sitting on the workbench behind me with wires dangling out of its guts uses a detector that our chemists tweak on a molecular level. I don't see how a technically-educated person who reads the journals and magazines could miss this.

P.S. The jet engine wasn't viable until advanced materials and high-bypass variable-stator designs were developed after the war. Even with the gung ho effort by the military industrial complex, modern jets didn't arrive until the late 1950s.

[Chris Phoenix, CRN]

Daniel:
Alternate history doesn't prove anything. But trying to learn from historical mistakes is worth doing.

The AI people were wrong about human-class AI being decomposable into simple sub-problems. Even language can't be decomposed, or we'd have a lot better machine translation than we do.

I said a nanofactory, once built, would build things that work by familiar engineering principles. You answered that the realm of the tiny is alien. Well, if you look closely enough, the realm of the macro is alien as well. That doesn't stop us from engineering in it. Just find a regime where things are predictable. Note that I cheated a little here: the building of a nanofactory assumes that such an engineer-able regime has been found. But it's not so much of a cheat: once we have even a few nanoscale devices that work, we can combine them at higher levels with *software* engineering practice. Keep them simple, put in enough buffers to control complex interactions (we'll have performance to burn), and the engineering stays tractable.

Nanoscale research is being done at a breakneck pace *in some fields.* I'm aware of that. But I'm also aware that in other fields that will be necessary to a mechanosynthesis/nanofactory approach, research is not being done. I don't see us working on anything that could make products competitive with even a basic nanofactory's output, except in a few narrow areas like computers. The only easy way to get the complexity into the product is with programmable fabrication. And the only way to scale programmable fabrication is to build small autoproductive systems. And we're not working on that nearly as much as we should be. Worse, we're going around saying it's impossible, so we're not having the policy discussions we should have.

Chris

[Flackbait]

Shelby-
While I agree that there are weapons better than knives and things more precise than nukes, I must say that I apparently did not make my point properly clear. Firstly, I was pointing out that there were competing technologies that, in important areas, do what nanotech simply cannot. For instance, the space-based weaponry-- ex. shaped orbital slugs. The practical progney of the British guided concrete 'bombs' used in Iraqi Freedom to take out enemy positions and armor with extremely limited collateral damage, the orbital slug has the potential to be as accurate as it would be deadly (granted, something moving at terminal velocity would still land with quite a shock, but that would depend on the mass of the slug). Short nano-production of a very precise laser mass driver that could catch the slug while it was still in the upper atmosphere and focus on it in an attempt to vaporize enough of its mass to divert its course, there's nothing nanotech could do against something like that.

Secondly, nanotechnology has a logistical problem-- even if it's nanotech, it's not magic. If you have a nano-filtration system, you still have to provide the raw materials and produce the things in sufficient quantities to equip your troops: something a nation such as China would inherently have trouble doing. If you have bio- or chem-weapon-neutralizing nano-agents, you still have to get them in a dispersable form and in large enough quantities to be effective. Ditto with nanotech-based WMD's. Logistically, wouldn't their greatest application be in simply producing better materials-- i.e. better armor, better projectiles?
And if that's the case, then realistically the US government doesn't need to be the one funding this. Shoudln't the private sector be able to come up with the proper nano-factories pumping out superior kevlar armor or whatnot without the ever-so-efficient government looking over their shoulders? If not, then the wrong side obviously won the Cold War, because the free market cannot look after itself.

[Flakbait]

Wow. All that, and I misspell my pseudonym, too. Aren't I great?

[AST]

This is an old and frustrating story. Even the things we do invent (LCDs, VCRs, etc.) get developed and made cheaply elsewhere, as our education system lags ever further behind.

Maybe we should leave nanotech up to Microsoft, that bastion of innovation.

I keep thinking that if the U.S.A. has the world's leading economy, the others must really be terrible. Maybe it's socialism.

[Bravo Romeo Delta]

I think one of the things that is being glossed over, particularly with reference to military development of MNT, is that the US has a much, much larger military than anyone else around. Right now, depending on how you count, the US accounts for half the world's defense spendind, with more than an R&D branch to match.

If we're not focusing on it now (and I'm not sure how much I believe the notion that we're not) it simply means that we'll pour money into it down the pike.

As far as a non-nano army winning a war, it rather depends on what level of nanotechnology the other guy has developed. Furthermore, nne mustn't ever fall into the trap of assuming that technology and technology alone wins wars - it is critical, as is logistics, training, communications, and so on. There are precious few instances when an army inferior in logistics, training, communications, leadership and so on beat another army simply by technology alone. One can look at any one of a number of Arab-Israeli conventional wars and take note of the fact that the Israelis had roughly comparable equipment and were at a numerical disadvantage to their opponents, yet managed wins.

This being said, a military as dominant as the current American military, equipped with full nanotechnological capabilities would be ferocious indeed.

[Engineer-Poet]

Daniel Newby must have forgotten the Korean War, in which jet fighters fought the battle for air superiority. True, that was not before VE and VJ day, but 5 years is a lot shorter than the 14 he implies.

Werner von Braun's group developed liquid-fuelled rockets starting on ground which had been trodden by Robert Goddard, an American. Goddard's work was ignored by our own government. Think for a moment what a few tens of millions spent on rockets could have done alongside the Manhattan Project; after WWII, the USA would have been able to push the Soviet Union back to its borders and prevented the Cold War.

Many people suffered and died because we missed those boats and had to play catch-up. Let's not do the same with nanotech; I'm tired of this story.

[RIch Rostrom]

Chris: re "early American airplanes were not very effective against the Germans..." When the U.S. entered the war, our main operational fighter was the P-40, which was quite comparable to the German fighters; and it was already being replaced by the P-38, P-47, and P-51. Our workhorse bombers, the B-17 and B-24, were already in service. All of these planes went through many improvements during the war as a result of experience, but there was no big R&D push to catch up to the Axis.

As for a "Nanhatten Project" - the Manhattan Project was directed at a very specific goal which was understood and known to be both possible and overwhelmingly powerful. No one in the nanotech world has put such a goal on the table yet.

[Chris Phoenix, CRN]

A-bomb known to be possible? There were people working on the project who believed up to the last minute that it wouldn't work.

Nanotech known to be powerful? A quick look at scaling laws and material properties tells you it'll beat today's products by many orders of magnitude.

Nanotech known to work: we at least know enough that we should be studying it a lot harder than we are. And we should be preparing for the likelihood that it does work.

A goal on the table? How about the tabletop production system outlined in Nanosystems twelve years ago?

Chris