Center for Responsible Nanotechnolgy (CRN)

Results of Our Ongoing Research

These pages, marked with GREEN headings, are published for comment and criticism. These are not our final findings; some of these opinions will probably change. LOG OF UPDATES

CRN Research: Overview of Current Findings

No Simple Solutions for Nanotechnology Risks

Overview: Molecular manufacturing creates several severe risks, and each risk tempts a simple and extreme solution. However, a patchwork of extreme solutions will be both destructive and ineffective. For example, Bill Joy and others have proposed halting nanotechnology research entirely. This would not actually work; instead, it would relocate the research to less responsible venues. The risks might be delayed by a few years, but would be far worse when they appeared because the technology would be even less controllable. To take another example, economic upheaval might be prevented by strict commercial licensing of all uses of the technology. This has two problems. First, digital protection schemes for commercial products have often proved quite easy to crack. Second, if the technology is so restricted that it cannot disrupt existing economic systems, continuing poverty will kill millions of people each year, fueling backlash, social unrest, espionage, and independent development. Each risk must be reduced by some means that does not exacerbate others. This will not be easy, and will require creative and sensitive solutions.

MM risks may include opposite extremes; so may attempted solutions.	Molecular manufacturing (MM) creates several severe risks, of several different types. For example, risks may be political, economic, or personal. Even within a single category, opposite situations may create risks. If incredibly cheap manufacturing drives down prices, economies may be disrupted. But if prices were held artificially high, the result would be concerted attempts to circumvent the restrictions—as well as widespread denial of humanitarian relief. With product cost potentially orders of magnitude lower than product value, prices may be unstable—they may fluctuate wildly as businesses abandon product lines and monopolies form, increase prices by ten times or more, and disappear under new competition. Or anticompetitive behavior may emerge; this seems quite likely under the circumstances, but this outcome would remove the benefits of free markets. In another category, allowing completely unrestricted use of MNT by everyone would provide powerful tools to criminals and terrorists; however, restricting MNT completely would require extremely harsh measures, constituting a risk in themselves.
Fighting one risk can increase another.	The problem gets worse when several risks are considered together. Attempts to prevent one risk, such as criminal use of nano-built products, may only increase another risk (maybe even in a different category), such as creation of a black market in unregulated MM. In addition, although many of the risks involve deliberate misuse of the technology, some occur even when all individual actors are behaving appropriately; such risks cannot be reduced by any simple strategy. CRN is focusing on the most dangerous risks—about a dozen of them, each of which needs its own handling, and all of which interact. Any MM administration program must work simultaneously to increase national security, increase economic security, and fulfill many other positive goals, while at the same time applying the necessary restrictions and policies to avoid instability in several different domains.
	For initial study, the risks will be divided into two broad categories: those that result from too much MM or too little regulation (the "permissive risks"), and those that result from too little MM or too much regulation (the "restrictive risks"). In general, attempts to avoid a permissive risk by increasing regulation will increase one or more of the restrictive risks, and vice versa. Any MM policy must also interact with the actions of users and developers/providers, both legitimate and illicit. Planning of a nanotechnology administration program must consider real-world motivations and situations, as well as multiple stages of cause and effect.
Nanotech must be restricted somewhat.	We can't afford to have unrestricted molecular manufacturing widely available to individuals. Criminals and terrorists would invent new products faster than society could compensate. Hobbyists would invent grey goo for the fun of it, just as computer viruses and worms are developed and spread today. Availability to individuals implies availability to governments, which implies arms races and various covert and nasty uses. Widespread individual use of molecular manufacturing systems would create at least as many problems as widespread use of personal computers: intellectual property violations, security problems... except that many of these problems would be translated into the physical world, where they could do far more damage. Even if MM is initially limited to governments and maybe a few corporations, it could create economic and political nightmares. Arms races and economic disruption would be quite likely. And as we will see, such limitation would increase the occurrence of restrictive risks.
Restrictions can be bypassed.	The initial developers of MM would surely want to keep control of it. How, then, could the technology become unrestricted? There are two pathways that must be dealt with separately. First, the technology can be stolen, cracked, "liberated," or otherwise accessed illicitly. Second, the existence of independent MM development programs greatly increases the chance of unfortunate use, and also provides more opportunities for technology theft. Two countries with independent MM capabilities could begin an arms race. Two companies with molecular manufacturing capability could begin a price war, resulting in either extremely low profits or extremely unstable prices. At least two of the risks, environmental damage from profligate MM use and social damage from undesired products, can occur even in the presence of some limitations on the technology. These risks can be prevented by severely restricting civilian use. However, as we will see, such restriction will probably be impossible.
Motivation for bypassing restrictions must be reduced.	There are several motivations that can lead either to independent development or to technology theft. From any nation's point of view, foreign attempts to develop MM may be perceived (with justification) as threats to the global balance of military or economic power. This will probably spark several crash MM programs. Once the technology is developed, any group without access to it will have a strong motivation to gain that access. National pride, intellectual curiosity, and simple rebellion can motivate attempts to own, develop, or control the technology for its own sake, or "crack" restrictions on it. If the use of the technology is costly due to licensing, financial interests may fuel attempts to acquire MM capability. If a disadvantaged group is unable to access it due to any form of restrictions—especially if the group is impoverished and MM could literally save lives—there will be a strong humanitarian and idealistic drive to make the technology available to them by licit or illicit means.
	These diverse and strong motivations, and the severe risks they lead to, invite a strong restrictive response. However, anti-nanotechnology regulation may be counterproductive. For example, any regulation that reduces humanitarian use of the technology will increase the idealistic motive. Any regulation that imposes severe penalties will increase curiosity and rebellion, while providing at best a partial deterrent. (Even the death penalty is insufficient to prevent some premeditated crimes.) Secrecy will likewise increase curiosity, and spark independent research. Any regulation that increases cost will increase the economic motive. It seems likely that a policy based entirely on restrictive regulation cannot long succeed. Many of the motivations can be blunted by making molecular manufacturing technology widely available under some more delicate form of control. Although this must be done with extreme care, it appears to be a necessary part of successful MNT administration.
Restrictions may themselves be risky.	As implied above, the dangers of unrestricted molecular manufacturing may inspire forceful and extreme restrictions, or even forceful preemptive actions to prevent MM proliferation. Extreme responses pose their own dangers. If carried to its logical conclusion, a policy of preventing any possible independent development would require worldwide restriction of technology to a pre-2000 level—perhaps even a pre-1950 level. Intensive surveillance might also be used to prevent any development, but the required degree of privacy intrusion (at least in the absence of near-AI levels of image processing) would be a clear violation of human rights. Thus the possible responses to MM risks create additional risks. Even in less extreme cases, attempts to preserve existing political or economic systems unchanged may cause the benefits of MM to be denied to those who need them most. Any MM-related disaster could increase the pressure for extreme restrictions. A technology leak, even without a disaster, could require (or at least inspire) very oppressive restrictions to avoid problems.
Restrictions may lead to a vicious cycle.	It is possible that a vicious cycle could develop between oppressive restrictions and attempts to break those restrictions. Criminals and terrorists are a small fraction of humanity with limited resources. But if MM technology is restricted enough to create prices that are obviously highly inflated, mainstream consumers will embrace any means to circumvent the restrictions (as they already do for entertainment). If the technology is restricted to the point that people are dying for lack of it—which is quite possible, given the number of potential lifesaving spinoffs—then even governments, humanitarians, and some of the technology controllers will have strong motivation to break the restrictions. Such widespread pressure to bypass the restrictions could inspire even tougher restrictions, costing more human lives and suffering. The likeliest outcomes are either widespread, long-term oppression, or an uncontrolled release of the technology.
Blanket restrictions won't work, but careful policy can.	The idea that control can best be maintained by giving up some control is counterintuitive at first, but it's simply a case of the well-known principle of diminishing returns. Any parent knows what happens if they try to micro-manage a teenager, and economists have long been aware that raising tax rates too high will result in lower total tax revenue. There is not enough space in a single article to analyze all the interactions of all the solutions to all the risks, but the examples given here are sufficient to show that in many cases, extreme solutions will backfire. Even if policy makers were willing to accept the humanitarian and economic losses of overly restrictive MM policy, such policy would be self defeating. However, this argument cannot justify a complete lack of restriction; the risks introduced by such a policy appear insupportable, and dealing with the problems would create harsher restrictions in the long run. Instead, policy must be sensitive to a wide variety of factors, including political, economic, and social dynamics.
	It should be emphasized that, although this page seems pessimistic, the larger picture is not so bleak. The very factors that make molecular manufacturing so dangerous—the rapid prototyping and unlimited manufacturing, and the immense complexity and power of the products—also provide unprecedented opportunities for positive outcomes. Even a small fraction of the raw capability would be sufficient to satisfy the world's humanitarian needs for generations to come. Another fraction could multiply the economy and enrich every owner of the technology. And only a small fraction of MM products are unacceptably dangerous. What is required is not blanket permissiveness or blanket restriction, but careful administration of each separate risk and benefit. It will take time to design and implement such administration, and it will be important in the nearer term to prepare for responsible administration by implementing responsible development.

DEVIL'S ADVOCATE —

You're expecting policy makers to be rational?

No—we're desperately hoping that they will be rational. Or at least self-interested, and willing to study the tradeoffs. Molecular manufacturing technology creates a large carrot and a large stick. CRN hopes to show the issues clearly enough to make it possible to avoid the wrong choices.

Next Page: Administration Options

Previous Page: Dangers of Molecular Manufacturing

Title Page: Overview of Current Findings