• 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

« Promoting Abundance | Main | Making Factories with DNA »

June 22, 2006


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

Nato Welch

I can see the usefulness of the model, for the most part.

Egalitarian democratic representation is a Guardian function. One might naively hope that a Commercial monopoly that begins infringing on Guardian functions would also take up democracy, in which case it might not be so bad. Alas, they never do.

I most critically distinguish Commercial from Government institutions by the way they distribute their decision making power. Commercial organizations do it one vote per dollar; Guardian organizations do it one vote per person. Other than that, they can be largely the same. I think this tendency toward egalitarianism is what makes Guardian organizations ideally suited to managing negative-sum transactions.

michael vassar

I don't think you "got it" at all Nato. Guardian organizations do not, in any meaningful sense work along the lines of "one vote per person", but rather according to heirarchies.

John B

Can anyone, anywhere, come up with a real-world example of any these systems *without* crossovers into other systems' behaviours?

I can't.

Take a look at the chart breakdown at http://www.crnano.org/systems.htm if you think you can - I'd be surprised if you could, honestly.

For instance, per that chart, only the "Information" system can use intelligence, only "Commercial" can trade, and only "Guardian" can defend.

IMO, these are *idealizations* of one way to split a philosophical hair. Jane Jacobs may have made that point in her work - I've not read it, and am only relying on the material posted here & at the CRNano website.

I do not think that debate based on these grounds is particularly helpful, since the reality doesn't follow this set of delineations very well.

If, instead of archetypes, you use this as the basis for something of a rating system, saying a given organization/system is a little "Information", a little "Guardian", but mostly "Commerce", that might be useful in some situations. Trying to decide nanotech policy based on such abstractions strikes me as building on sand, however.

John B

Chris Phoenix, CRN

John, I agree that the real world is never as neat as the categories we make. A policeman walking a beat may stop into a store for a stick of gum: a commercial transaction.

The point of Three Systems is that the crossovers are potential points of weakness/corruption. And an organization that tries to straddle between two Systems is going to have to work very hard to avoid corruption. Look at modern media, or modern academia. Or the modern government/industrial complex, for that matter.

From my point of view, the fundamental insight/argument is that:
1) Zero-sum (e.g. security), positive-sum (e.g. market), and unlimited-sum (e.g. information) transactions are fundamentally different;
2) Managing and optimizing these different types of transaction will require dissimilar approaches (this is demonstrated/supported by the very different value systems);
3) A single organization that tries to live by multiple value systems and address multiple dissimilar problems is likely to become unethical and destructive;
4) MM will create zero-, positive-, and unlimited-sum issues, and these will have to be handled by a non-simple, non-unified approach.

The abstractions here are pretty fundamental--not based on observations of historical organizational structure--that's just supporting data, and of course it'll be real-world messy.


John B

It could be much worse than your initial example of the policeman and the stick of gum - the alleged & unwritten 'quota system' in many American communities in which police finance themselves by writing at least a certain amount of tickets per month is one possibility. (See http://www.thenewspaper.com/news/05/587.asp for a quick google hit along these lines.) Truely a Guardian style system using Commerce. This admittedly shows that crossovers between the styles are potential points of corruption.

However, a guardian system that dissimilarly guards certain aspects of its organization - the existance of guardian systems more than the right of privacy of the people which make it up, perhaps - are other points of possible corruption. (Examples of this kind of behavior are left as an exercize to the reader.) If all you're saying is that the models indicate potential points of corruption, you're offering a flawed tool which doesn't anticipate the full range of corruption.

This still leads me to question the value of debate along these lines.

Further, given your statements in your first paragraph of this reply ("the real world is never as neat as the categories we make"), the gist of your second paragraph ("the crossovers are potential points of weakness/corruption"), and your third point ("A single organization that tries to live by multiple value systems and address multiple dissimilar problems is likely to become unethical and destructive") indicates to me that your logic-based position is to expect all organizations to have to work very hard to avoid becoming unethical and destructive.

Not arguing with what I take to be your conclusions - I just derive 'em via different routes.

-John B

Chris Phoenix, CRN

John, "Not arguing with what I take to be your conclusions - I just derive 'em via different routes."

If my conclusions are derivable via different routes, that implies that both our routes are more likely to be valid.

I expect all organizations to have to spend some energy on not becoming unethical. As they do (with a few exceptions like Mafia). For most orgs, their incentive is that they'll be punished for violating the structure of laws/government. For governments, well, democracy works to some extent, as long as people are willing to work for constructive/ethical courses of action rather than ignoring looming problems.

Interesting point about Guardian orgs treating some of their charges preferentially. I'd tentatively compare this to dishonest sales people in commercial orgs. Not necessarily a problem with the fundamental architecture of the org, but rather problem that a well-designed org will try to minimze.


John B

Chris -

This could be misread, but I'm sorry, I just do not agree that the ends justify the means. By which, even if we're coming to the same conclusion, the reasons behind our reaching these conclusions makes a major difference across the board of discussion. (To be clear, I'm not accusing you or CRN of anything. Merely disagreeing with your statement above.)

My arguement to reach a similar endstate relies on the skullduggery I see inherent in humanity - self-interest & laziness being IMO two of the biggest, if not the top two, motivators of bulk humanity (yes, including myself). In short, I'm a pessimist to your optimistic take on society & people.

This affects a lot more than just this one position. I believe it affects the probability of an effective world government or organization to regulate nanofactory templates. It affects the likelihood that even if nanofacs are produced, they may or may not be 'free' or 'cheap' to own and use. (See the legendary(?) story of JP Morgan & Tesla, where the latter offered to power the world and the former asked how it'd make him money. Allegedly Tesla coudln't think of a way, and the idea wasn't funded.)

I agree that all organizations will need to to spend effort to avoid "becoming unethical." I think that an open society is more likely to be able to bring internal & external force to bear against transgressors of the accepted ethical/legal system, be this boycotts or regulation or other techniques, "guardian" or otherwise. I do not dispute that democracy works to some extent. *wry grin* However, I am reminded of Churchill's great quote - "It has been said that democracy is the worst form of government except all the others that have been tried."

Interesting that we've now got examples of Guardian internal corruption and Commercial internal corruption. Have to see if I (or if someone else beats me to it, chime in!) can come up with an analogue in an Information system. I would be deeply shocked if there weren't.


John B

Just bumping this back onto the 'recent' list, curious about Chris's response (if any)


Chris Phoenix, CRN

To some extent, the answer justifies the method: a method that picks a right answer out of a lot of wrong ones is likely to be valid. And if two, very different, methods come to the same right answer, that's very interesting.

But it sounds like our agreement was less specific than I thought. Simply deciding that institutions are likely to engage in skullduggery is not a very innovative answer.

It's worth exploring institutional sources of corruption. I mostly agree with your view of human nature, though I don't like it. But in addition to individual corruption, there are also ways to set up institutions that are more or less corrupt. Leaders can set the tone, expecting people either to follow the rules or to break them. And organizational structure and focus, I argue, can also affect the ease with which it can become corrupt.

Context can lead to corruption. For example, the vice squad of a police department walks a very narrow line, and its members often slip over. Conflicting imperatives can lead to corruption. Were the police who refused to let the Katrina survivors cross the bridge to safety corrupt? Why or why not, and by whose standards? Perhaps the individual police were not corrupt, but the organization was?

If conflicting imperatives can lead to corruption, which I expect we'll agree on... and if each of the Three Systems has its own conflicting imperatives... then an organization that tries to bridge the gap between the systems is likely to be corrupt.

I realize that I'm largely just restating my position. But I think I'm getting a bit narrower and more nuanced each time, while staying with the core insights. I'm hoping to find a statement that you'll agree is interesting and fruitful, to identify a (not the only, but a major) source of corruption.


John B

Chris -

Sorry if I wasn't clear there, and I wasn't. *wry grin* See if I can do better...

"It's worth exploring institutional sources of corruption" - Agreed. As you weren't clear that you were only looking at one subset of the problem where these systems met up, however, I felt it necessary to cry 'foul'. If you were to state that this is useful in its limited scope, I'd agree it's one of several limited POV's which have utility in their area. And I agree wholeheartedly that this is useful to "identify a (not the only, but a major) source of corruption".

"I mostly agree with your view of human nature, though I don't like it" - Neither do I. Unfortunately this is all we got to work with and within, tho'. *shrug* I do find that agreement somewhat surprising, however, given how strongly optimistic CRNano's positions have been regarding many of the social issues surrounding nanotechnology/nanofacture.

I would agree that efforts along the nanoblock venue have a stronger security stance than many which have been mentioned thus far elsewhere. I'd agree that you're approaching issues others either ignored, assumed 'trivial', or otherwise didn't deal with.

But it seems to me that your chosen solutions - an effective world control of nanotechnology by some organization or body, among others - are extremely optimistic, especially given that so much of the basic work underlying the concepts is public domain. It'd be one thing, if it were all wrapped up in a black lab or labs somewhere with strict security - but it's not.

-John B

Chris Phoenix, CRN

John, that's our *proposed* solution, not our *chosen* solution. We don't know enough yet to choose a solution. (Sorry if that sounds like a cop-out; we're working on it.)

I don't think we're expecting human nature to be the factor that makes nano institutions good. Rather, it'll take good design of the institutions, lots of accountability and openness, channeling self-interest into positive applications, etc. The US government lasted over 200 years, changing leadership every few years, with only a few constitutional crises, so it is not hopeless to think of designing a good or at least self-limiting system out of flawed and greedy parts.

Your comment on basic work being public domain makes me think: Why do we have only one Internet? We could have lots of Internets, but we have one integrated one because it's a lot more efficient and valuable that way. Perhaps with MM, a technical infrastructure that allows things to work together smoothly will be valuable enough that even weapons will be built using it. That could go a long way toward accountability. Just an idea...


John B

Sorry, that strikes me as weasel-wording - you 'chose' to 'propose' a personal nanofactory solution. *shrug* If you'd gone with another possible solution as the center of your models of how nanotech would evolve over time, you'd have chosen to propose that. I agree, we don't know enough yet to say what the final solution(s) will be. *shrug* This is one reason I try to caveat my discussions here and elsewhere to fit in with the general concensus of the site in question - *assuming* a given convention, what happens?

Regarding human nature vs institutions, we disagree on some of the fundamental assumptions. You're assuming that people will work on such efforts with the best goal they can conceive of as their target. I agree that far, but I disagree that their goals are going to be purely altruistic and instead are likely to be selfish. IMO, many of those in positions of power, capable of making policy decisions affecting the developmental path(s) of nanotech, see this as akin to Star Trek or other fantasy. This would indicate to me that they have little to lose (from their perspective) making the process do 'good things' for them. This is perfectly rational from their point of view, and there's little we can do to show 'em the error of their ways. YET. (Even if we can show 'em lots of doodads, I rather doubt all of them will catch the importance of what nanofactories will mean...)

Unfortunately, there's lots of politics around nano already, and as funding increases with few of the more earth-shaking results which have been promised (at least thus far), nano research is threatening to become akin to AI research - a deathknell for a career as people who fund such research find they feel they've been mislead or duped by promises of those actually performing the experiments.

*shrug* I did warn you that I'm a pessimist. *sour grin*

Why do we have only one internet? Good question! Because of the way it was developed, IMO. It started out small, and grew over time as technology improved and new uses were found for the technology. This contrasts with your comment that "it's a lot more efficient and valuable that way", in that it grew incrementally and still has non-trivial amounts of inefficiency and value-eating costs associated with it. There was no grand vision of what we accept as the internet today back in the earliest days of hooking computers together - no one seems to have predicted personal computers, for instance, nor the vast changes such have brought about over the last 20-odd years.

Note that this is rather different from the modern story of "atomic-precision nanofacture" (rather than nanomaterials like some sunscreens or chips), which IMO is growing in spite of not having any serious sucesses to point at, just a whole list of maybe/might be/could be/should be scenarios.


Chris Phoenix, CRN

John, there's a difference between "suggest" and "command." (And as Heinlein said, between "roger" and "wilco.") What I meant is that we haven't settled on it as THE solution, but have proposed it as A possible solution.

On human nature and institutions: I'm not even sure that people will always act in their own self-interest. What I've been hoping--and it may well be naive--is that if an institution is confronted with a problem that threatens to destroy it and kill all its members, that institution will make necessary changes. It seems very clear to me that MM is such a problem.

On the Internet: There was not originally one Internet. There was BITNET, for example. There were lots of private BBS's. And remember the ! syntax for email? But the advantages of interoperability eventually made everyone use TCP/IP and sendmail.

On nano researchers: If they're really lucky, MM will develop fast enough that they'll be able to jump on that bandwagon and claim to have been working toward it all along. They'll explain that MM is really hard and needed all sorts of background research first, but they really always supported it... really sorry that our technology didn't turn out to be on the bootstrap pathway (not that it ever had a chance of being relevant, in most cases)... but please don't think your money was wasted, because look, here comes MM, rah rah rah!

BTW, what would you consider a serious success for atomic-precision manufacture at this point?


John B

Chris -
Were CRNano focussed in any way on solutions which involved other possible nanotechnological bits beyond personal nanofactories, I'd have less trouble agreeing with you. However, from my POV it seems quite clear that the organizational focus is pretty tight on personal nanofactories. For instance, your recent "Printable Robots" post goes on to specifically incorporate the idea of how much more effective a personal nanofactory could be.

I agree about people not always acting in their own best interest - the so-called "rational actor" is IMO a myth or an archetype, not a common occurrance. Unfortunately, while "the prospect of being hung concentrates the mind wonderfully" it doesn't necessarily concentrate it in ways which increase probability of survival. We're rationalizing beasts rather more than rational ones, IMO.

Agreed on the Internet, the "network of networks". It has evolved - in as messy and painful a manner as biology exhibits - into the 'single network' we know today through many iterations, trials, and attempts. The main advantage is that technical evolution (such as the Internet has experienced/demonstrated) does not rely on reproductive rates, but rather on fiscal rates and the creativity of those attempting to increase the profits to be had from such technology. (These profits don't have to be monetary, just an improvement in fulfilling some need.)

My point being, it's not been a single goal from the get-go, which was how I read your comment about its being so "because it's a lot more efficient and valuable that way."

*chuckle* True, success has many parents - at least, after the fact.

Some possible serious successes for atomic-precision manufacture would include demonstrable, reproducible solutions for one or more of Dr. Jones' "Six Challenges". Another includes a well-designed series of tools, something along the hydrocarbon metabolism paper, which covers the whole range of reactions needed to create the full tool set & other objects. ('Well-designed' includes the fun calculations needed to ensure stability in the expected environment, error detection/remediation, etc)

The latter is effectively the home run, as it gives the immediate path to at least a subset of atomic precision nanofacture. That by itself may be the requirement to get enough people to sit up and pay attention to the problems.

*shrug* Hope this helps,


Chris Phoenix, CRN

The internet is the way it is because that's valuable, in the same way as the cheetah is the way it is because that's survivable. Neither one was planned.

On personal nanofactories, I'm a little confused. Earlier you said, "your chosen solutions - an effective world control of nanotechnology by some organization or body, among others..." Now you say, "Were CRNano focussed in any way on solutions which involved other possible nanotechnological bits beyond personal nanofactories, I'd have less trouble agreeing with you. However, from my POV it seems quite clear that the organizational focus is pretty tight on personal nanofactories."

Do you see PN's as one of the "others" of our chosen solutions that your first quote referred to?

Most of our references to PN's are made, not becaue they're a cornerstone of our proposed policy, but because they're a convenient way to understand some of the technical possibilities and implications of MM. Basically, PN's are a useful way to boil down our message to something comprehensible.

We have suggested that widely-available limited PN's might be a good thing for the world, spreading benefits while reducing demand for unlimited PN's. We've basically recommend this, in the absence of analysis that demonstrates it's a bad idea. The lmiitation might or might not be implemented by a global governmental body. It might be a commercial standards organization like UL.

I agree the well-designed series of tools would move MM forward. Freitas proposed to produce that, plus some synthesis sequences for machine components, for just $5M.

Jones's six challenges are misleadingly stated. To take just one example: In challenge 2, he asks, "Will it be possible to engineer complex mechanisms in the face of this lack of dimensional tolerance?" But really, about the only thing that requires extremely tight tolerance over a distance is mechanosynthesis. The same softness that allows diamond to wobble also allows machines to accommodate that wobbling.

I've now read the entire web page you pointed to, the challenges and all the comments. None of the challenges worries me. They are all stated in ways that exaggerate their apparent severity, which makes it difficult for me to know just how much to discount them. But as far as I can tell, the fundamental issues that they bring up can all be engineered around. One example: To purify a fed-in molecule stream, it's not necessary to make a sorting rotor that rejects all water. It's sufficient to build intermediate chambers with *exit* rotors that have a high affinity for water. Just a few such chambers in series can remove water to ridiculously small concentrations--too small even to disturb a nanofactory.

If this solution works, have I solved Jones's challenge? No, because it is not my solution. Drexler described it in 1992 in Nanosystems. So why was it not addressed in the challenge?

It is things like this that make me not take the challenges seriously enough to address them all in detail. I just don't think it's a good use of time. For example, I have a dim memory of the friction paper he cites, as perhaps focusing on much higher speeds than are contemplated in Nanosystems. But I won't bother to check it because I've checked many of Jones's claims in the past and have always found that they did not in fact demonstrate the problem they were supposed to demonstrate.

Eventually I got tired of researching problems that turned out to be non-problems, especially when I started to notice that a little bit of thought on Jones's part should have shown that they were non-problems. Either Jones missed some obvious points (and I think he's smarter than that), or Jones knew about them and presented the problems anyway (which is good debating, but very bad science).

The other reason I don't often read what Jones writes is that debating while pretending to engage in science *really* pisses me off, and I tend to write overly blunt opinions that people for some reason blame me for expressing. For that reason, I do not plan to read Jones's response to this. This post is bad enough--I don't want to be tempted to write a followup.


John B

"The internet is the way it is because that's valuable, in the same way as the cheetah is the way it is because that's survivable. Neither one was planned." - Is that why the appendix is here? Because it's valuable? How about personality theft? *wry grin* While the theives definitely make money in personality theft situations, IMO it's NOT an evolutionary net-positive...

Sorry for the confusion re: what I see CRNano promoting. I was cutting too many corners. Here's my thumbnail sketch of my understanding of your suggestions:

World-wide control & enforcement of production designs for personal nanofactories. There would be some form of testing/validation of production designs for safety. Once accepted, the designs are available to all personal nanofactories, either on a fee basis or as 'freeware'. Personal nanofactories would accept a specified feedstock, data (the production designs) and power and produce zero-chemical-pollution products and heat.

In short, I see both a world-wide controlling organization with enforcement ability and personal nanofactories being keynotes to your current-best approach. Your comment above that you use 'em because they're "a useful way to boil down our message to something comprehensible" is quite understandable - the whole thing is complex as hell on all SORTS of levels! However, I might suggest going into excruciating detail on both the benefits and negatives of the situation, specifically addressing your reasoning. (My apologies if you've already done this - I've not run across it in your sites thus far - unless you intend the 30-questions framework?)

The link to Drs Freitas & Merkle's goal are IMO a great idea that hasn't yet happened, and with the recent Foresight org's retargetting, their mission's been "superceded", to use the wording at the top of the page you linked to.

Regarding Dr Jones' challenges, I personally don't think you answered it. I agree that as a broad-brush approach, that's a 70%+ solution, but there's still quite a bit of additional effort which would need be addressed. Filtration such as you suggest helps remove the great bulk of the problem(s), but it doesn't guarantee the perfect feed of feedstock which I understand mechanochemistry to require. (Using mechanochemistry for the moment because biomimesis doesn't have this problem)

You've already done a lot of high-level discussion on the other <30% of the problem - specifically, endproducts are tested. Those which are failures are recycled (somehow). Given that very high purities CAN be expected in such filtration schemes, it could be argued that you'll be able to (with the additon of a few additional layers of filtration, worst case) reach purity levels high enough that you're producing an arbitrarily high ratio of successful products to those which fail testing.

Dr. Jones IMO (you'd have to ask him for his opinion) wouldn't accept this as a valid respose, as the solution level you're looking at doesn't address many of the issues he considers important. (Sort of like your confusion about why I'm so picky on words, from elsewhere on CRNano, his reasoning seems to me to indicate he's got reasons for his pickiness. He'd call it attention to detail, however.)

I'm sorry to see you & Dr Jones argue, honestly. You're both bright, dedicated people, and I can't help but wonder what benefits you might both gain by working together rather than maintaining disagreement. *shrug*


Chris Phoenix, CRN

'' "The internet is the way it is because that's valuable, in the same way as the cheetah is the way it is because that's survivable. Neither one was planned." - Is that why the appendix is here? Because it's valuable? ''

Um, we were originally talking about one attribute of the Net, not the whole thing. "The way it is" referred to that one attribute. More wasted time.

On the Jones challenge, I wasn't trying to give complete answers, just to demonstrate that he is making the problems sound worse than they are. I don't know why he wants to do that. (Perhaps then he can get more credit for helping to solve them.)

I was answering specifically the point about small molecules sneaking past sorting rotors, and I think I did give a complete answer--or as complete as it can be at this stage of engineering, *and* complete enough to demostrate that the problem is solvable.

Solvable does not mean 100% perfect feedstock--there is no such thing. Mechanochemistry does not require 100% perfect feedstock. As long as the feedstock has a lower error rate than that imposed by background radiation, I'll be happy. And that is probably relatively straightforward to achieve, for small-molecule feedstock. In fact, I think it's safe to say that a cubic meter of feedstock can be reliably purified of all contaminants--except for radiation-induced transformations of feedstock molecules, of course. (I'm also thinking that error correction is a separate topic from filtration/purification. I can solve 100% of the problem of excluding small molecules to less than 1 part in 10^20, and then solve the separate problem of dealing with 1 error in 10^15.

On arguing with Dr. Jones: I spent a lot of time talking with Dr. Jones. I talked with him even when it became apparent that he was setting exercises that he knew the answer to. (The way a professor treats a student, not the way a researcher treats another researcher.) I stopped when he made a misleading claim in a public conversation in order to win a debating point. By ceasing to engage him, I *stopped* arguing with him.

I think our early discussions were productive, as when we agreed that reversibility and efficiency were closely related. And his promotion of floppy systems led to my insight about entropic springs in proteins--which ironically demonstrates that floppy systems are fundamentally lower-performance than machine systems.

The current style of the challenges indicates a moving and artificial target. If challenged on any point of them, past experience suggests that Jones would engage me in interminable debate (not productive discussion) in which the problem to be solved constantly shifted. I am not maintaining disagreement; rather, if we talked, we would be maintaining disagreement (debate).

If Dr. Jones had published a version of the challenges in which I could not identify missing solutions that I'm sure he already knows about, then I would be more willing to suggest solutions that he may not have thought of yet. Meanwhile, I consider the challenges misleading and would like to spend as little time as possible on them.


The comments to this entry are closed.