• 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

« The nano-Rorschach | Main | Assessing Nanofactory Capability »

June 13, 2004


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

Chris Phoenix, CRN

The restrictions on the above-linked page are not a recommendation; they are a list of mostly-independent possibilities. Policy makers will pick and choose from those plus many others. At first sight, they appear to be technically possible and to increase control without hurting efficiency too much.

Isn't it a cop-out to list a bunch of restrictions and then say we don't like some of them but we won't tell you which ones? I don't think so. Thing is, we really don't know which ones are good ideas. We don't have a policy worked out.

We have some general guidelines, and we're not too sure about those. One of them is: don't massively destabilize the geopolitical system. So, Mike is more or less right that we're trying to reduce disruption of the status quo. Where we disagree is that we think this is very relevant to safety. People tend to get miserable and/or die when political systems are destabilized.

About space flight: We don't talk about that, simply because we don't know what to say yet. Is universal access to space consistent with keeping any degree of control over molecular manufacturing technology? Probably not. Is that sufficient reason to give up all control and create a nano-anarchy? Personally, I don't think so, at least for the first few years. I just don't think nano-anarchy is survivable, at least for the people left on earth.

So, given a choice between 1) immediate access to space, with a few thousand people taking advantage of it, and a few billion getting hurt; or 2) delaying access by a few years, in which we learn to live with this amazingly powerful technology, and then we can all go homestead comets or whatever; I'd choose the second. But are these the choices? I have almost no clue.


Brett Bellmore

Of course, the very way you describe the options assumes that your regulatory regime succeeds. Keeping your eggs in one basket IS convenient, we'll all agree, in the event the basket doesn't get dropped. If it does, you've got utter disaster.

I'd urge you to consider what happens if we follow your chosen course, don't disperse mankind through the solar system, and your regime doesn't work. The worst case scenario involves the extinction of mankind...

Perhaps we ought to consider how we can colonize space in a big way, WITHOUT compromising your at home regulatory scheme. "A few years", you say. And, how long does it take to get to other bodies in the solar system, beyond the Earth-Moon system, using minimum energy trajectories?

A few years. You're essentially demanding that we not colonize space, to preserve a system YOU don't expect to last as long as the first colonization trips could easily take!

So, lauch the ships, on long trajectories, with limited resources such that they can't turn around until they've reached their destinations. By the time anything could return, it's all shaken out for better or worse on the Earth.

At least, accept that barring space colonization just fuels the pressure to overturn your regime. It's only chance to survive in the first place is to either rest lightly on people, giving smart people as little reason to evade it as possible. OR for it to become an unbreakable tyranny. Your control philosophy keeps you drifting towards the latter pole, it seems to me...

Chris Phoenix, CRN

The minimum-energy trajectory idea is clever, but won't work. Any self-sufficient colony ship would have the resources to send back a small probe.

All our eggs will probably be in one basket for a while no matter what we do. And then, unless an utter disaster happens (and I'd consider unbreakable tyranny an utter disaster), the eggs will disperse. The time interval is less important than the events. Do we want a global panic/disruption leading to war? A few terrorist acts and spooky accidents leading to (attempts at) renunciation or lockdown of the technology? Obviously not...

We realize that any non-disastrous scenario in a nano-enabled world will be fairly brittle. From the start, we've proposed policies aimed at minimizing the desire to overturn it and maximizing the buy-in from a wide variety of power blocs.

I'd like nothing better than for someone else to propose a workable plan for preventing the massive problems that MNT could create. Go ahead and try it. We've established that our proposals will face opposition and will be difficult to get right. I'll find it much easier to abandon them if someone can come up with an idea that seems at all likely to work. (Saving a small sample of humanity and letting the earth go to hell doesn't count.)


Mike Deering

Even if these decisions will inevitably be made by politicians, the questions of how serious are the dangers and how restrictive do the controls need to be to avert them are scientific.   To list a range of restrictions and present them to the leaders of the world to pick and choose from is tantamount to asking them, "How much power do you want?" which is a political not scientific question.   The people at the pinnacles of politics got there because they are drawn to power.   They can't be trusted to make this kind of decision.   Every president since George Washington has sought to increase his power and to remain in power as long as possible.   Is there ever a question of a seated president running for a second term?   No.   Your intentions may be as pure as you describe but the result will be the most complete totalitarian regime ever conceived.   And even then, your objectives of containment and security will fail, and it will all come to a bad end.

The geopolitical status quo can't be preserved and is not worth preserving.   It can't be preserved because the technology is already too diversified, politically, geographically, and culturally.   And it can't be preserved because the technology is universally underestimated, even by those who have it, therefore sufficient measures will not be taken to contain it, even the most restrictive of yours.   If it can't be preserved, then attempts in that direction will be counterproductive in the extreme.   The geopolitical status quo is not worth preserving because it ignores the suffering of millions of innocents.   Forty thousand children die every day from preventable causes; wars, not having vaccinations, dirty wells, and lack of food.   That's nearly fifteen million a year.   If the governments of the world spent ten percent of the money devoted to guns and bombs and soldiers on solving these problems they could be eliminated today with current technology.   Why are we in this situation?   Because, to the leaders of the world, these children are not a priority.   These are the same people you want to have pick from a list of restrictions that range from mild to total, and give them the opportunity to become gods over us all.

Of course, if you frame the question as an alternative between extinction and totalitarianism, well then, your approach starts to look a little more attractive.   But that line of reasoning is based on a very high estimation of the likelihood of the worst dangers becoming real.   "How serious are the dangers?" ,is a scientific question.   Ignoring gray goo for obvious reasons, how dangerous would a nanotech arms race be?   You claim, very dangerous, and justify the extreme restrictions on your list.   I claim, not very dangerous, and justify a looser set of restrictions that completely avoids centralized control of the technology, but not nano-anarchy.   These are my scientific reasons to support my claim:

1. In the age of nanotechnology, arms quality not arms quantity will determine the victor.   How smart, targetable, versatile, adaptive, controllable, are your weapons is much more important than how many tons of medium tech weapons can you produce.

2. The leader in innovation, creativity, genius, science, and engineering has always been the more free society, free to exchange information, free to use and develop technology, free to control their individual destiny.

3. If we go with a centralized control of the technology we lose this advantage, and we lose the war.   Game over dude!

I see a nano arms race as inevitable and the only options are: we win the race by going with our current strength, our personal freedom, or, we go the route of totalitarianism through centralized control of the technology and we lose the race.

What do you see?

Mike Deering

"I'd like nothing better than for someone else to propose a workable plan for preventing the massive problems that MNT could create. Go ahead and try it." says Chris as he throws down the gauntlet.

"Challenge me will you?" asks Mike as he picks it up, serendipitously slipping a brick into it, and lets him have it back across the right cheek.

Okay, here's my plan.

Shipping costs are negligible. The nanofactory is the size of a mustard seed and self-assembles, upon being dropped into a gallon of gasoline, into its functional desktop form. It produces, from commonly available materials, 262144 copies of itself (in seed form) per hour. Mailings are scheduled to arrive from April 18, 2005 through April 28, 2005 inclusive. Mailing arrival dates are scheduled according to geopolitical considerations. Areas lacking resources and with poor and underdeveloped assess to technology first. Resource rich and high-tech countries last. Multiple modes of delivery will be utilized. We will need your mailing address by January 17, 2005. The seed contains a sophisticated computer, all the information required to produce the nanofactory, and all the machinery required including several different types of nanobots. Upon activation, the nanofactory connects with a wireless world-wide information network. Some questions you may want to ask can not be answered at this time due to considerations of compartmentalized security.

The whole New World Order thing is one way to go with nanotech safety. As usual, there are two sides to this issue and I think they both have some good points. On the one hand, in order for this technology to generate the most good for the most people it needs to be widely available, meaning everyone should have their own desktop nanofactory. On the other hand, if everyone has the capability to whip up weapons of mass destruction, none of us will be safe. And you can easily find people advocating both ends of the spectrum of control, from no access, to unlimited unregulated total access. One extreme throws the baby out with the bath water and the other doesn't have good long term sustainability. If neither extreme is good, then we enter the real world and the question becomes; how do we integrate our need for security with our desire for personal access to the technology?

I think Kurzweil makes a good point that open source active defenses personally and institutionally deployed are the most effective safety measures available. I would combine this with a multi-tiered system of controls including three different kinds of nanofactories.

Consumer product nanofactories: unrestricted use of "safe" nanofactories with tamper-proof built-in limits against producing weapons. These nanofactories could not produce any biologically active or living organism such as a virus or bacteria, or biologically active molecule such as ricin or other poison, or any high-explosive or high energy material such as C4 or nitroglycerin. It could produce any electronic devices or mechanical or structural devices or materials, also food stuffs made from an approved list of safe nutrients, and over-the-counter medications. These nanofactories would be free to everyone.

Industrial Nanofactories with tamper-proof built-in software security capable of producing only products from approved certified designs. These nanofactories would check the security code of encrypted design files submitted to them prior to producing each product. Product designs would be approved and encrypted by government certified facilities. Users would have to pass a security clearance screening and submit to government monitoring. These nanofactories would be owned and operated by companies producing potentially dangerous products such as industrial chemicals, explosives, pharmaceuticals, and biologically active organisms.

Research Nanofactories housed in government controlled secured facilities for government use in national security, law enforcement, and research. In addition to government uses, private researchers could come to these facilities to do research too dangerous to do elsewhere.

In addition to these controls I would like to see proactive government enforcement of laws against the misuse of nanotechnology by individuals or nations, including severe penalties.

I think it would be very difficult to produce gray goo due to design considerations, and the biological threat of nanotechnology can be addressed while still allowing the masses to reap the benefits of personal nano-assembly technology with a proper balance of liberty and controls. I think the top level of controls needs to be on a world-wide scale. Preserving our present state of multiple independent sovereign nations is non-viable. Whether this means a world-wide alliance with the US in the lead or the US taking over the world is about trivial.

Most people agree that widely available nanofacs capable of producing WMD's would be a significant treat to security. What we want our nanofacs to be able to produce are food, housing, vehicles, and toys. This unfortunately requires some restrictions be placed on the nanofac functionality. The nanofacs we are distributing are classified as consumer product nanofacs.

It is very easy to make nanofacs impossible to tamper with therefore there is no danger of the nanofac being hacked, or converted to other uses. It merely requires that all functional elements of the nanofac be implemented at the molecular level and be distributed in a 3d web-like network throughout the diamondoid structure. No one without molecular manipulative capability would be able to disassemble or modify the device, and anyone with molecular manipulative capability would not need to hack the device, they could make their own from scratch. This is a good reason not to make the internal design of the nanofac modular, as yours is.

The way these restrictions will be implemented is that the nanofac will be a two stage device internally. The first stage produces a variety of design approved nanoblocks including diamondoid structural components, diamondoid mechanical components, molecular electronic and logic components, and a variety of nutritional food components. These nanoblocks will be not less than 200 to 500 nanometers in smallest dimension and contain no nano-assembly capability. Additionally there will be pre-approved designs for larger nanofactories, medical nano devices and nano security devices included in the first stage. The first stage is all hardwired, non-programmable, non-modifiable. The second stage can assemble the nanoblocks into any product or finish the assembly of the hardwired approved designs. When the second stage is operating in the nanoblock assembly mode it is fully programmable, in the hardwired assembly mode it is non-programmable and under the control of the first stage.

The nanofac will have its own wireless network connection to other nanofacs and the internet. It will serve as your communications hub for your wireless telephone, TV, computer, radio, or any other communications device. The nanofac will be able to produce a wide variety of common consumer products and foods. Files for these produces can be downloaded from the internet. Additionally, you can design your own products from any combination of approved nanoblocks, and produce your product, or post your design on the internet for anyone else to use. The reason this is safe is because WMD's require fissionable materials, or high energy explosive materials, or nano-assembly capability, or live organisms. The list of approved nanoblocks will not include any fissionable material, high energy material, nano-assembly capability, or living components. Things you can not make with the nanofac include C4, nuclear weapons, chemical weapons, viruses or bacteria, animals or people. Things you can make with the nanofac include but are not limited to: big macs, very powerful computers, TVs, cell phones, furniture, houses, cars, planes, submarines, solar cells, clothes, robots, and a limited number and style of larger nanofacs, nano-medical devices, and nano-security devices.

jim moore

I share your assessment of and concern with the current geopolitical system. In another post I will write about how MNT might be used to peacefully transform that system.

But unfortunately, I don't share your optimism about an MNT arms race.
1.) With MNT if you get there first and you are clever and genocidal, you could just kill all of your potential enemies. (*note* I don't want that to happen.)
2) Quality of strategy can be as important and sometimes more important than the quality of the weapons.

The only thing that really gives me hope about full scale nano-warfare is that it is such an obliviously bad outcome for everyone. And the alternative, were we all get along peacefully, is such an obliviously good outcome that we will choose peace over war.

Mike Deering

You know what? I was cutting and pasting some things from other documents in my last post and some stuff got in there that wasn't supposed to. Please disregard any references to mailings and dates.

jim moore

That was obviously bad and obviously good not obliviously bad and obliviously good.

It obliviously time for me to go to bed ;-)

Janessa Ravenwood

Mike, you’re still disregarding 2 things:

A) We’re almost certain to get Broadcast Architecture nanofacs first. It may be some time before we figure out seeds. What do we do with all of those BA nanofacs? Leave them sitting around to gather dust until we figure out seeds?
B) ANYTHING can be hacked. Your statement of “impossible to tamper with” is extremely funny from my perspective as a programmer. You just don’t GET the hacker mentality (not to mention their track record of pretty much 100% success eventually). The more you lock it down, the more of a technical challenge it is to hack it and the greater the prestige of the hacker who pulls it off. Eventually, one of them will. Also, as both Brett and I have pointed out, you can use nanofacs to build the tools to build the tools to build the tools to bootstrap your own. This will also eventually be done as well. So WHEN the “tamper-proof” nanofacs are cracked and someone spams the world with the blueprints to bootstrap their own totally unrestricted nanofac (or eventually drops one of your seeds in the mail, but as an unrestricted version), what then?

Brett Bellmore

As for the idea that you could prevent the manufacture of chemical weapons by denying the end user nanochemistry control... Do we manufacture them that way now? I'd imagine an obvious early product would be a programable MEMS chemistry lab; It would consist of inert chambers and pipes, heaters, refrigerators, pumps and valves, all certain to be standard blocks. In goes water, air, dirt. Load the right software, and out comes anything from diet soda to Sarin gas.

Mike Deering

Janessa, I'm trying not to forget anything but thanks for pointing these out.

A) That other simpler nanofac designs will come out first. I'm expecting A.I. to make significant improvements to our nanofac designs even before we produce the first one.

B) Anything can be hacked. If you don't have access to the hardware on a molecular level, and you don't have access to the software, you can't hack. The operating system can be in hardware. Ever tried to hack the micro-code of a processor chip? Ever tried to hack your own DNA without nanotechnology?

The product design and editing software can be run on a PC with windows, but the nanofac won't accept any commands that violate its security protocols.

Janessa Ravenwood

A) And if that doesn't happen as quickly as you anticipate?
B) Once AGAIN, we can build the tools to build the tools to build the tools we need using the available nanofacs. You keep assuming that's it's an ABSOLUTE that those tools will forever be out of the reach of your average hacker. I have no idea why you believe this.
C) No security protocols are hack-proof. Which might be the easier solution to all of this - THAT part is software and NO security software thus far has proven totally hack-proof.

Janessa Ravenwood

To further clarify, you say the security protocols are to be hard-wired? Excellent! That makes the job easier. Ever downloaded a Windows security patch? When your copy of Windows leaves the factory it’s static. If a vulnerability is identified in it, that vulnerability remains unless you install new system files to correct those parts of the OS that are vulnerable to attack. Your proposal means that any vulnerability CANNOT be corrected once identified. Woo-hoo! Oh, you say patches will be available to fix identified flaws? Even better! The security protocols now are NOT fully hard-wired and false patches can be installed to help take over. Also, you can de-compile the patches to help understand the security system better (which is what some enterprising Windows hackers have been doing). So which is it to be? Either way has vulnerabilities to exploit.

Janessa Ravenwood

Even further clarification: It’s easier to hack the computers of yesterday than the computers of today. Your current system proving hard to hack? No problem. I’m immortal in a mature nanotech society so time is on my side. I’ll stuff a few copies of the current machine in my basement and wait about 15 years (10 Moore’s Law cycles). I’ll THEN use current technology and go to work on those old machines. I’m willing to bet I eventually crack one of them. Then I’ve got all the nanotools I need. Who cares if they’re out of date. With them, I and like-minded individuals can catch up to the current state of the art as now we certainly CAN build the nano-scale tools to hack the current nanofacs. So what then?

Janessa Ravenwood

Yet another issue on (A) - what if some company starts selling BA nanofacs anyway? Can you control all companies on the planet in every country? The required nanoscale tools for nano-hacking will likely be available then. What then?

Sorry for so many posts, I seem to be on a run of hacker inspiration today.

Brett Bellmore

To be fair, Windows is probably a bad example to use when discussing the possible security of a hypothetical system; Windows security is so bad I've often suspected that Gates writes viruses for a hobby... Mike probably isn't proposing to ship the nanofactories with their security features turned off, and give the user the option of activating them if it happens to cross their minds. LOL

Given the use of terabyte one-time pads to communicate insanely long keys, which are used for one transaction and then thrown away, communications between the nanofactory and some central server could indeed be secure. (Then there's quantum "cryptography" to consider, if you rebuild the communications infrastructure.) And the factory could be built to be "paranoid", and capable of self destructing at the first hint that somebody was trying to compromise it.

No, I think the most likely security failure would be at the source, through covertly inserted back doors and cheat codes. That, or the simplest failure of all: No monopoly to begin with.

Janessa Ravenwood

Which I very highly suspect will be the case - see my last post in the series.

Janessa Ravenwood

Home nanotechnology labs!

Via Instapundit, this is the best news I've had all day! And this is in 2004. Can't wait to see I can personally afford in 2014. So much for building a nanotech monopoly...


jim moore

I think we need to approach security with a layered and multifaceted philosophy.
I see these as the facets:

a.) We need to make it difficult (but not impossible) to do "bad" things with MNT.
b.) We want to make sure that we reduce to a minimum any incentives to do something "bad".
c.) We need to have a system in place make sure that people who do "bad" things are stopped or get caught.
d.) We need to have a rapid response system in place when "bad" things start to happen.

In general I see "bad" things as being destructive or coercive or non-consensual acts.

Chris Phoenix, CRN

Mike: We're not posting the control options for governments to see. We're posting them for citizens to see, and realize that governments will think of them, and want to use them, whether we post them or not! If you're outraged, good! But don't shoot the messenger (that would be us). Figure out some way to avoid those options being 1) attractive and 2) available to 3) governments; 4) without disastrous nano-anarchy.

Janessa: I'm sure I've said it before, but I'll say it again: we don't expect these techniques to keep limits on nanotech for even fifteen years. They'll either be abandoned, or replaced by far more draconian techniques. I don't think we can predict more than 5-10 years after MNT, if that. MNT products, including AI platforms, will advance far too quickly.

Mike, again: Your plan actually doesn't sound that different from ours. What do you see as the major differences? BTW, I'll shoot one hole in it: your product-filtering scheme is way too simplistic. Rail guns and flywheel bombs can do as much damage as any chemical explosive. This kind of thing is why I think products will have to be checked case-by-case for adequate security.


Janessa Ravenwood

Chris: I'm already satisfied that you know that no restrictions will last forever, but Mike doesn't seem to think so, hence my arguments were directed towards him.

Though I will ask you - and I don't recall receiving an answer from you on this - what happens WHEN unrestricted nanofacs become available (from a regulatory point of view)? You're so geared towards establishing this monopoly that I think you're disregarding what to do after it's lost or (far, FAR more likely) never comes about in the first place.

BTW, I was also thinking that Mike's scheme was starting to sound a lot like yours so I'm wondering what his fuss is all about - "Your scheme of strictly regulating nanofacs is intolerable! So instead I propose...strictly regulating nanofacs!" At least that's the way it's looking to me.

michael vassar

I think that the plans for what to do when unrestricted MNT is available can plausibly fall into two camps.
1) Flee into space (this won't save you)
2) Self-enhance your mind to come up with a better idea (if done in a hurried manner this will probably lead to destroying much of what you value about yourself)

Mike Deering

Granted, a terrorists can kill a lot of people with high tech conventional weapons produced by a nanofac, but he can't wipe out all of humanity or all life on Earth. A national army equipped with high quality conventional weapons can cause a lot of trouble, but it can't prevail against a worldwide alliance of nations that have advanced active nanotech defense systems.

We need to be clear on what our objectives are here. Do you want a regulatory system that would prevent your crackpot neighbor from killing you and everyone in your neighborhood? Sorry, that would require a system that would be too restrictive. Or do you want a regulatory system that would prevent any terrorists organization or rogue nation from destroying a large country or even all life on the planet? That we can agree on. Stop trying to regulate rail guns and flywheel bombs and stick to engineered viruses, offensive nanobots, and rogue A.I.'s. Use conventional methods (law enforcement) and nanotech tools to deal with conventional threats. Use technical countermeasures (active blue goo, nanofac design limitations, SAI's) to deal with radically new classes of weapons and threats.

Don't worry too much about not being able to maintain the monopoly on nanotech. The defensive capabilities, when implemented in a properly multi-tiered system, equal or exceed the offensive capabilities.

The difference between CRN's plan and my plan is what level of restrictions are recognized as necessary.

CRN's plan:

Limits on production of conventional hardware (rail guns, etc).
Checking for approval from Central Authority before every production cycle.
Self destruct sequence activated by loss of communication with Central Authority.
Centrally produced nanoblocks or purified chemicals required for input.

My plan:

No limits on production of conventional hardware (rail guns, etc).
No checking for approval from Central Authority before every production cycle.
No self destruct sequence activated by loss of communication with Central Authority.
No centrally produced nanoblocks or purified chemicals required for input. Runs on grass clippings.

Now can you see the difference?

Janessa Ravenwood

Mike: thanks for the side by side clarification, helps much. Yes, that does sound much better. However, I wonder that if we have some central authority certifying nanofacs and their restrictions then - in practice - it might start to resemble CRN's plan.

One other thing - you've probably got a better chance of restricting the crackpot neighbor than the terrorists or rogue states. Frequently because the rogue states shelter the terrorists. And if you're not going to go to war against the rogue states to stop them, well then they're not going to be stopped. See North Korea, Iran, Syria, Saudi Arabia, etc.

Mike Deering

Janessa, Bush's war on terror is a separate issue. Even after the Singularity, whichever regulatory system is implemented, it may be necessary to be proactive about the enemies of civilization. I don't have anything against war, providing nobody gets hurt.

The comments to this entry are closed.