• Google
    This Blog Web

October 2011

Sun Mon Tue Wed Thu Fri Sat
2 3 4 5 6 7 8
9 10 11 12 13 14 15
16 17 18 19 20 21 22
23 24 25 26 27 28 29
30 31          

RSS Feed

Bookmark and Share

Email Feed

  • Powered by FeedBlitz

« Will personal nanofactories ever be built? | Main | Molecular manufacturing and wealth concentration »

September 16, 2006


Feed You can follow this conversation by subscribing to the comment feed for this post.

Nato Welch

When you say "almost no one", I tend to think the rub comes from the term "almost".

We regularly recognize the empowering capability of MM, the effect of democratizing more power into more hands. With this technology, each individual can have a greatly magnified effect in comparison to the effects we can have without.

If that's true, then that tiny minority you refer to with "almost no one" will have their effects on the rest of society greatly magnified with MM. Those effects will more easily reach the majority who may not be so creative or ambitious, effectly having such change imposed upon them, be it for good or ill.

"Never doubt that a small group of thoughtful, committed citizens can change the world. Indeed, it's the only thing that ever has." --Margaret Mead

Michael Anissimov

"But how many people would actually go to the effort and discomfort of changing their lives, just so they could use this new toy to its fullest potential? When the question is asked that way, the answer seems to be: almost no one."

I disagree strongly... many people LOVE to "push the limits". You are being too conservative here, in your projections about what people would do. There will always be "hackers" who want to push their hardware to the extreme - and this is more like one out of every fifty people rather than one out of every million, as you seem to be insinuating.

I personally know people who would probably just let their nanofactory print out endless crap all day long, just because it's amusing. Millions of cameras to spy on people, millions of little homunculi for a mini-army, millions of solar cells to provide power, millions of habitation cells for the purpose of making a huge pseudo-mansion, millions of bits of garbage just for the sake of having a footprint. Maybe this is something that teens will enjoy more than adults, but it's definitely something that many people will certainly do.

A good way to minimize social disruption from nanofactory-level technology will be to LIMIT ITS CAPABILITIES as much as possible. Why on earth do you think that by maximizing the capabilities, we can minimize the footprint? Some kind of bizzare reverse-psychology? This is just silly.

All it will take is one front-page story of how someone did something cool by pushing the limits, and then everyone will try to copy him.

Michael Deering

Myspace is a good example of the creativity of the general public. There are 110 million people with myspace pages. How many of them have ever created any original content? Most of them have copied content from elsewhere on the web, pictures, text, html scripts. The greatest number of original content is the easiest to produce, blogs. Half of them don't even have blogs. How many of them have created original graphics for their page? 1 in ten thousand would be my guess. Of course there are lots pf websites producing content specifically for myspace that is completely free. Talk about reverse IP protections: they want you to copy it and put it on your page.
Designing a product is likely to be much more complicated than designing web page content. So who's gonna do it and give it away like myspace content? Coca-Cola will give away a file to produce a frig that reminds you to by Coke. Media companies will give away a files produce toys that remind kids to watch the upcoming episodes of cartoons. Lexmark will give away files for printers that only use their certified ink cartridges. Guess what Cingular will be giving away?

michael vassar

Michael A: People who work with SIAI shouldn't start sentences with "I personally know people who" in the course of ANY discussion of the distribution of normal human behavior.
Identifying all the cognitive biases contained in your post is left as an excercise for the student.

Chris: I don't think that terribly many people need to use their nanofactories to change life fairly totally. If businesses adopt them, many many people will be unemployed. If healthcare providers adopt them and advertise, health will dramatically improve (and it only has to be relatively unregulated in a few countries). If governments even think other governments might adopt them we have unstable arms races.
Also, even if "normal" people don't do much with MNT at first, some less normal people will, and the quality of lifestyles that those people will be able to create will be appealing to other people and will attract new people to join them.

And, of course, there's the usual issue.

Chris Phoenix, CRN

Michael, I said maximize convenience, not maximize power. Release the fabber with a lot of cute-but-useless designs. Things that don't really hurt if someone prints out a million of them. And remember that it doesn't directly use molecular feedstock, it uses nanoblocks purchased from elsewhere; you won't be able to make a million things cheaply.

Eventually, other designs will arrive. But by that time, most people will have gotten tired of the new toy, and may never bother to search for the new designs.

And there's room for social engineering even in the illicit designs. Make "3D porn" available before advanced weaponry, and a large number of the illicit-design-seekers won't bother to look for the weaponry.

There's also room for engineering in the nanoblock mix. Put in motor-blocks at a rate of only a kilowatt per kilogram. Someone wanting to do more extreme products will probably want a more "active" mix of nanoblocks. Sell it to them... and track it.

As to media coverage of extreme acts, do you think the media is that irresponsible/free? I don't.

I just tried to google for stories or instructions on the use of gasoline as an amateur explosive. After several searches, I finally found one explosion, done by a group of artists. Why don't these one-in-fifty extremophile people ever try to make big combustion bangs, and then talk about it? Maybe a few do... (and maybe the FBI knows about it and shuts them down...) but it's apparently not in the memespace. There's just not a lot of point. Flammable liquids are too mundane.

What I'm arguing is that if nanofactories can be made as mundane as gasoline right from the start, their destructive potential will be as underused as gasoline's. At least by the general public. The military, of course, will make napalm.


Chris Phoenix, CRN

Too many Michaels: my previous comment was addressed to Michael Anissimov, not Michael Vassar.

Michael V, I'm not trying to solve all problems of MM with this suggestion. Just trying to minimize the potential for sudden massive private misuse when the first nanofactories hit the private market. Economies and militaries are a whole 'nother question.


Eric Tufts

I for one tend to agree with Chris but to a lesser degree and perhaps for a slightly different reason. And to be fair to Chris, he's only postulating... having us imagine "if". The reality is that nano-factories will initially appear with some strings attached. It's almost certain that whoever introduces the first nano-factory, be it a startup with large amounts of venture capital behind it or a major corporation with shareholders to satisfy, will require a business model that protects it's own interests and the interests of it's stakeholders. I've been thinking about this problem for the last four years and wrote a business model about 3 years ago that closely matches what Chris described in his posting. That is, exponentially manufacture the blocks separately then sell them as feed-stock for the nano-factory. It's very much the existing printer/ink business model; sell the printer at low or no profit then charge for the ink with massive profit margins. Those ink profit margins are in fact what kept HP afloat during the PC industry downturn early in the decade.

It won't be a revolution as such for several reasons. The cost of the nano-factory item might provide only marginal savings over store bought items once you factor in feed-stock and IP. The beneficial impact MM could have to help those most in need would be negligible pending goodwill or legislation. Another factor is that shopping has become a social behavior more than just a necessity and social behavior is very difficult to change in a short period of time. Even if everybody had a nano-factory that probably wouldn't stop most people from going to the mall and making those impulse buys (research on this aspect would be interesting and valuable). Not all companies will initially provide their IP to be available to the nano manufacturing process as they take a wait and see approach. We can look to the iTunes Music store as an example of this as it took Apple years to sign labels and networks on. And as Chris mentioned, how many people are going to be designing their own products? Sure, once the software becomes intuitive enough for the average user, but software has always lagged behind hardware and it could be years before that becomes a reality. There will be a group of people, perhaps in the thousands or even millions that are early adapters and will take to designing their own stuff much as there are millions of software developers that develop for a paycheck or open source for the greater good. An entire economy could develop around these more personalized designs and could compete directly and successfully with established brands. There will likely be government regulations requiring approval on new designs which can then be added to a universal library. Nano-factories would then only be able to download designs from this approved library.

The good news concerning these apparent drawbacks is that MM is likely to emerge slowly, allowing the economy to adapt as necessary and making massive unemployment overnight less likely. How long the economy needs to adapt is of course anybody's guess. Hopefully, by the time the first nano-factory company starts to put traditional manufacturers out of business, other nano-factory companies would arise to compete with it, and as competition increased, prices would drop. The hardest hit economies will likely be manufacturing based economies such as China. Manufacturing accounts for only 13% of the U.S. economy, and many of those items (cars, planes, military equipment, large household items, food, etc.) won't be feasible for home nano-factories to manufacture at first. As people slowly lose their jobs in manufacturing other jobs are likely to appear (such as designing products). it does seem there will have to be some level of involvement from the government to help make the transition more fluid especially with higher education/job training.

In the longer term we could see more capacity from home nano-factories such as non-confined (cars and homes) and mechanosynthesis (food) leading to as Chris put it "retirement". So grab as much land as possible before then.

Dan S

1. If nanoblocks are priced high, how long would it take for unrestricted nanofactory (capable of use raw, cheap materials to produce blocks) to "leak" to general public? Given that one can made as many copies of it as he wish, providing he had one.
2. Flammable liquids are simple wrong model for malicious nanofactory products. Look at computer viruses. Dozens of new viruses are produced each weak. There are many virus source codes ("designs") available on the Net and there are virus generators you can use to make new virus by pushing a few buttons.
3. Few creative users doesn’t mean no one, and I think even a single "creative" user can do much harm. And there are creative users out there who “like to push the limits” (look at processor overclockers for example – how about 1,83 GHz Core 2 Duo running at 5GHz?)
4. It is current economic model that minimizes creativity. Our system needs many people with little creativity otherwise it can not function. In the long run, post-MNT economy has potential to change this. And how long really would it take?

Phillip Huggan

"1. If nanoblocks are priced high, how long would it take for unrestricted nanofactory (capable of use raw, cheap materials to produce blocks) to "leak" to general public? Given that one can made as many copies of it as he wish, providing he had one."

The idea is that a nanoblock factory won't consist of the components necessary to build products from, say, molecules of acetylene. The molecules of acetylene are 100-10000X (I'm not exactly sure) smaller than is a single nanoblock. Other vital components to a diamondoid nanofactory such as the tool-tip and some actuator components, are much smaller than will be the size of the blocks. If someone can make actuator parts and tool-tips from a nanoblock contruct, the nanoblocks are no safer than is a mature nanofactory.

Tom Craver

Dan S:

(1) A big part of the nanoblock fabber concept is to greatly reduce demand for atom-precise fabbers. If, in addition, it is highly illegal to own atom-precise fabbers, very few would be willing to take a chance on it.

Actually the inverse problem is more likely - atom-precise fabbers might leak out before nanoblock fabbers can be designed and released. Taking away something that's already leaked out - even if it was always illegal to own them - will be tougher.

(2) It seem most computer viruses aren't very destructive - they're more like self-spreading graffiti.

(3) Agreed - but Chris' flammable liquids analogy does seem to hold here. One person with a dozen Molotov cocktails could probably start a hugely destructive fire - but most people don't consider that a "creative" idea, and so don't do it.

(4) I hope MNT does increase creativity, and help make people more self-sufficient. The more that happens, the less I think we need fear that people will become alienated by the loss of their old lives, and do something evil.

If I can get a new job installing a vast solar and wind power array, or spend my laid-off time eliminating my electric bill and setting up a backyard automated greenhouse to cut my grocery bill, or start a new fabber-enabled business, I'm going to feel better about my ability to survive and support my family.

Dan S

I understand we have two kinds of devices then: a nanoblock assembler (highly guarded) and nanoblock-enabled nanofactory (avaliable to general public). I would argue that this is not going to work either:
1. How many nanoblock producers will exist? One in the world? One for a country? Or may be several for every city?
2. If there are many nanoblock assemblers a lot of people would have access to them. Suppose someone will use it to make few diamond mechanosynthesis nanoblocks and then use them with "conventional" nanofactory.
3. Now he can make any number of unrestricted nanofactories.

Dan S

"One person with a dozen Molotov cocktails could probably start a hugely destructive fire - but most people don't consider that a "creative" idea, and so don't do it."

One person with nanofactory could probably do more then "start a hugely destructive fire". This person may for example first use it to develop unrestricted nanofactory (there are many ways really) and then setup his own private empire somethere (a lot of options again). Even if stopped somethere in the way, such person can cause a lot of trouble in the process.

Phillip Huggan

"2. If there are many nanoblock assemblers a lot of people would have access to them. Suppose someone will use it to make few diamond mechanosynthesis nanoblocks and then use them with "conventional" nanofactory."

Diamond mechanosynthesis is hard will be hard. For instance a single deposition reaction still hasn't been accomplished yet despite Nanosystems being in print for almost 15 years... obviously whoever supplies nanoblocks won't supply an SPM-like nanoblock capable of mechanosynthesis.

"This person may for example first use it to develop unrestricted nanofactory (there are many ways really)"

There may be no way to develop an unrestricted nanofactory. For instance, everytime you use mechanosynthetic depositions to try to make a necessary diamond gear, the teeth may reconstruct either immediately or during operation, to form a useless disk. Nanoblock manipulation has probably been doable for over a decade using semiconductor fab tools that can be purchased with no restrictions. Diamond mechanosynthesis is technically a much more difficult problem.

Michael Anissimov

Michael Vassar,

I do personally know typical people, so it would be ridiculous for me to pretend that I don't.

Also, I know *atypical* people who would print out millions of random nanoproducts, just for fun, if they had the ability to do so.

Not too many biases in my post... I'm just arguing that people push the limits. I've followed up on the dangers on unrestricted nanofactories at my blog:


Tom Craver

Dan S: "If there are many nanoblock assemblers a lot of people would have access to them. Suppose someone will use it to make few diamond mechanosynthesis nanoblocks and then use them with "conventional" nanofactory."

One of the ideas in "Safer Molecular Manufacturing through Nanoblocks" was that the devices that produce nanoblocks might be "fixed function" - i.e. have the program to make nanoblocks hardwired in at the "atoms" level, where it'd be extremely difficult to modify without destroying the device.

So there would likely be many nanoblock production units, probably sited near large energy sources, in licensed facilities with a reasonably high level of security.

Those fairly common units would be produced in turn by a few hard-wired units, which in turn would be produced by a very few fully programmable atom-precise nanofactories. The latter would be kept in a few facilities with most-extreme security measures.

Dan S

Diamond mechanosynthesis is hard today. However, if one is capable of printing 10 petaflop computer (or even 10 or 100 such machines) he can use it to quickly find a plausible pathway to diamond (or silicate-based or boron nitride) mechanosyntesis. And then print or buy necessary equipment (SPM’s for direct approach or biolab for biopolymer pathway).
Actually if there are any pathways from proteins to mechanosythesis then equipment commercially available today should be more than sufficient. For direct approach one probably needed better SPM than available today but again we simple don’t know what pathways could be found using vast computing resources.
So, for restricted nanofactory approach to work, you must put heavy restrictions on what can be produced (definitely no billion CPU’s laptops!) and this will make nanofactory just a toy with no more capabilities when LEGO. Also put SPM and standard biochemistry equipment under heavy control. Actually, stop any future progress in this area, since future developments could easily lead to unexpected pathways. This is no different from simple permanent ban on all developments and applications of advanced molecular nanotechnology. And consequences of this are well discussed on CRN main web-site.

Phillip Huggan

A problem with overplaying just how easy diamond MNT will be to achieve experimentally is that it makes brainstorming MNT defenses very difficult and it draws ridicule from the experimental SPM community who know just how damn hard a target a single diamond mechanosynthetic operation will be to achieve.

Computer simulations are vital in determining what mechanosynthetic depositions are viable experimental targets and will be vital in determining which mechanosynthetic products are stable. Computer simulations will tell us just how powerful a technology diamond MNT will be and also if other diamondoid substrates are worthy of closer inspection. But simulations don't provide any clues for *how* to build tool-tips, how to functionalize them, how to optimally construct a feedstock reservoir. These are the realm of engineering and chemistry.

I can't comment too specifically here (because I'm no chemist or surface scientist), but the alphabet soup of nanotechnology processes and tools that will be used to prepare feedstock reservoirs and deposition sites, along with fabricating and functionalizing tool-tips, owe little if anything to computer simulations. Simulations tell you what, but not how. A computer AI expert system would need to be a better chemist or diamond surface than is a person, for brute force to help you here.
Maybe you are right Dan, and these currently unknown (and possible impossible) techniques will be common in a decade or two along with the tools. But I'm looking at progress now, and it is slow (by nanotech standards).

Chris Phoenix, CRN

Phillip, I largely agree...

But it would be a huge help if we could clarify exactly what physical capabilities an SPM will need, and be pretty sure of what machine we needed to buy and how to program it.

Freitas's work today takes 50,000 hours of CPU time to write a paper on one reaction--out of potentially hundreds (counting all the different lattice-building chemistries). If that were reduced by a factor of 1000, then any hobbyist could do work that's currently being done only one or two places in the world. If it were reduced by 100,000, then new kinds of investigation could take place, such as automated searches that today would not be feasible at all.

Yes, simulations aren't the same as lab work. But easy simulations 1) get a lot more brains working on the problem; 2) make the lab work significantly more likely to succeed.

It's true that cheap simulations won't invent a nanofactory. But they may let a very smart tinkerer invent a nanofactory.

They will have another effect that must not be underestimated. Currently, it takes a polymath genius to really understand nanoscale mechanics. But if people can build stuff and watch it work in simulation, they will be able to develop physically reasonable intuitions simply by looking, rather than by thinking very hard about equations. Again, simulations will greatly increase the brainpower that can be directed to the problems.


Phillip Huggan

Yep. And even non-DFT simulations like Nanorex's, have the effect of channeling more minds to MNT-relevant university courses.

Tom Craver

Dan S:

The point of making nanoblock fabbers isn't to create an absolute barrier to anyone ever making an atom-precise nanofactory. It's a multi-dimensional attack on the problem, including:

- It satisfies the bulk of demand - without this, vast numbers of people would lust to have their own atom-precise nanofactory, creating huge economic demand for someone (possibly backed by criminals) to re-invent it.

- Authorities, having atom-precise MNT, can create devices and systems that are somewhat more efficient and capable than things most people can build, allowing them a better chance at countering any threats created using nanoblock technology.

- It limits the availability of atom-precise component designs. So if someone does develop their own atom-precise nanofactory, they'll have to invent their own components, before they can use it for evil ends. Authorities will have a huge "lead" on anyone who re-invents MNT, giving them an edge in dealing with evil uses of MNT.

- There'd be a program providing controlled and monitored access to atom-precise MNT - greatly reducing the 'curiousity' factor for those bright enough to develop their own illegal nanofactory, and making them "insiders", which will tend to make them empathize with and side with authorities.

- As there is so little legitimate reason to possess or independently re-invent atom-precise MNT, anyone doing it can be assumed to be up to no good - so those would be illegal. The illegality alone will disuade many people from experimenting independently. Those who are not disuaded may be more easily detected and arrested than would be the case if everyone had a form of atom-precise MNT.

None of those things *prevent* someone from re-inventing MNT. They do greatly reduce the probability that someone will want to, or be willing to risk it, or (if they do anyhow) that they will risk using it in large ways.

The best solution to governments re-inventing MNT, is to simply let every "reasonable" government in on it via a treaty organization, and use the power of that organization to suppress MNT activities of "unreasonable" governments.

Of course, they'll have to sustain the will to do so, and not back down when some wacky-evil government decides it needs nanoweapons.

Tom Craver

One reason I so dislike the inkjet printer cartridge model for nanoblocks:

Making nanoblocks artificially expensive runs diametrically in opposition to the reason for limiting people to nanoblock fabbers in the first place.

It creates pent-up demand for a personal nanoblock-maker, which needs to be atom-precise, creating demand for criminals to fund development of illegal nanofactories.

It would also virtually guarantee that nanoblock manufacturers won't make them recyclable, in order to maintain their revenue stream at a high level - further exacerbating the cost issue.

Nato Welch

@Dan S:

There would be nothing preventing hackers from building compatible "pirate" nanoblocks interoperable with existing ones, yes. But in order to be effective, they'd either 1) need to have a supply chain in place to distribute the new blocks, since existing fabs wouldn't be capable of building them, or 2) they'd have to build and release a separate recursive fabricator that can make these "pirate" blocks. Since the ability to build atomically-precise nanoblocks is far over and above the common, public fab units, it's just as easy to build & deploy a fully-fledged atomically-precise nanofab instead.

That being the case, the hackers wouldn't be exploiting any vulnerabilities in the design of the nanoblock/microfab design - the point of the design isn't to prevent people from devising full-fledged recursive nanofactories (and hell, what could?), but to //reduce the motivation// for people to do so. In the sense that the nanoblocks architecture enables as many beneficial applications as its equipment is capable of without engendering user resentment through arbitrary restriction of the system's native capabilities (ala DRM), its pretty likely to succeed //at the task it's designed to accomplish//.

Those trying to hack the microfab/nanoblock architecture will reap fewer rewards than they will from hacking a DRM'ed atomically-precise nanofactory, and that will reduce the motivation to do so.

Furthermore, those wishing to enable novel but merely antisocial applications (eg, copyright circumvention devices) will be able to do so without opening up the opportunity for far more dangerous applications (eg explosives, grey goo). Hacking a general-purpose restricted nanofactory presents a "class break" that a microfab/nanoblock system doesn't. Once one person publishes vulnerabilities in a DRM system, the uses broken nanofacs can be put to are extremely broad. With nanoblocks, legitimate researchers can limit their published vulnerabilities to mitigate potential damage much more easily, since exploiting the vulnerabilities will likely depend on the availability of nanoblocks made by equipment not found in the wild, and which legitimate researchers are not likely to release widely.


I need to find away to hack my electric power with out opening it or is there something that i can build to slow it down.

Tom Craver

" it's just as easy to build & deploy a fully-fledged atomically-precise nanofab instead"

Precisely - by creating an artificial scarcity of nanoblocks - in order to maintain profits - the value of nanoblock-based nanofactories in de-motivating attempts at making atom-precision nanofactories is greatly reduced.

Jahlove - you could install solar power panels and feed power back into the grid - that literally does what you want, AND it'd be legal (if properly approved and installed), unlike what I suspect you're asking for.

The comments to this entry are closed.