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

Finding a Solution that Balances Many Interests

This page is more preliminary than the others. CRN is not recommending these solutions yet; we need to do more research before we will know what can work in the real world. However, we do think that each of the problems addressed by these solutions must be dealt with somehow.

Overview: The Center for Responsible Nanotechnology has developed a tentative outline for the international administration of molecular nanotechnology (MNT). Under this proposal, a self-contained, secure molecular manufacturing system—a personal nanofactory—would be developed in a closely guarded crash program. The personal nanofactory (PN) would be released for widespread use. A PN would only be able to make approved products, or approved classes of products, and the approval process could be quite flexible without giving up too much control. Very few products, even military products, require a built-in molecular manufacturing capability. Families of products could be classified according to increasing product safety and MNT containment. Only unusually dangerous products would require any human approval. At the same time, the built-in restriction infrastructure would allow military, commercial, and societal interests to be protected. Intellectual property could be protected without discouraging innovation or preventing humanitarian aid.

A nanotech program must balance many benefits and risks.	Any good molecular nanotechnology administration program will promote at least four benefits, while preventing about a dozen major risks. This is not easy, and is proof that a simplistic solution cannot work. The program should promote personal, institutional, and international security; economic well-being of groups and individuals; humanitarian relief and human rights; and innovation. In addition to the risks previously analyzed, the existence of a global MNT administration implies the possibility of corruption, so the program must be carefully designed to minimize that. A program that fulfills all these requirements should be acceptable to most groups and people; this is good because it will need cooperation from almost everyone to prevent the inevitable few malicious people from bypassing it.
It's safest to develop one nanofactory quickly.	A key to our proposal is the use of a flexible, carefully controlled manufacturing system. As described in our Technical Restrictions page, self-contained molecular manufacturing systems (personal nanofactories) can have restrictions built in to prevent many kinds of misuse, by allowing only approved products or classes of products to be built. This allows a fine degree of control over the whole technology—as long as the restrictions are not broken, and as long as no competing development programs exist. Competing programs are a bad idea for at least two reasons. First, they multiply the chance of a technology leak. Second, they reduce accountability and trackability, which increases uncertainty and decreases security.
	The nanofactory project should begin as quickly as possible, both to preempt competing programs and because of the millions of lives that could be saved each year by marvelously inexpensive molecular manufacturing. The project should use the best of government, industry, academia, and open source talent and resources. The more people and groups who contribute in some way, the fewer will be working outside the system on their own MNT projects, and the less chance there will be of independent projects reaching the goal first. However, some parts of the project should be highly secure. Much of the difficulty in building an assembler comes from the lack of a reliable recipe; release of all project data would probably make a rogue project too easy. The idea is to create "the only game in town"—the only project that has a realistic chance of early success—so that everyone feels the need to support it.
Products can be approved by a flexible process.	Foundational to this proposal is the idea that no special interest group can be allowed to restrict the technology to the point that it is not useful to other groups. Nanofactories should be widely available—one in every store, one in every village, possibly one in every home—to maximize the benefits and profits. Many different groups can have input into the product approval process. The security group must be able to veto any design that would allow a technology leak. Governmental jurisdictions will want to prevent undesired products, and different governments will have different ideas about what is undesirable. Support of intellectual property rights will require the forbidding of designs that violate trademark, copyright, or patent. Conversely, in order to maximize the utility and benefits of the technology, most designs must be approved quickly.
	There are at least four levels of risk. The most severe is technology leaks leading to unrestricted nanofactories. Next is massively destructive products. Next is dangerous products. Finally come products that are illegal but not destructive. Whole classes of design can be approved as "probably safe". Products in these classes may be illegal or even dangerous, but undesired designs can be dealt with even after a few copies are produced. In more dangerous classes, designers and builders, and the products themselves, may need to be licensed and tracked. A product that released microscopic diamond fragments, or that used a lot of power, would probably need more careful review, just as many products today are UL listed or CE certified. Still, the "probably safe" classes provide much scope for innovation. Most of the products used by most people today, and most of the early humanitarian products, would be considered "probably safe".
	A single, adaptable nanofactory with a flexible and efficient product approval/disapproval process appears to satisfy the requirements for usability. A well-designed process can satisfy at least five groups simultaneously: military, businesspeople, humanitarians, users, and innovators. Each of these groups has very different goals, methods, and outputs. The next few paragraphs will consider them individually.
Well-administered nanotech can increase national security.	Nations may attack each other to improve their situation (e.g. by seizing a resource), to remove a threat, or because of bad leadership. Nations maintain militaries both to deter and resist attacks from others and to prepare to attack others. (In some nations, the military also provides internal policing.) MNT can provide almost any physical resource, reducing one incentive to attack. Molecular manufacturing can relieve all desperate domestic conditions caused by lack of resources, and can even make a big dent in conditions caused by lack of education. However, unrestricted MNT could increase the perceived threat from other nations. Nanotech weapons, developed and deployed in secret, could be quite destabilizing. If neither side knows what the enemy may develop or how to counter it, they may be tempted to launch a preemptive strike when they believe they have a momentary advantage.
	The main question, then, is how countries that do not want war can be secure in a world with MNT. The ideal situation is one in which everyone knows that no attack can succeed. This requires some level of knowledge of each other's defenses—which each nation should be happy to advertise as a deterrent—and some level of knowledge of offensive capabilities—which they may not be happy to advertise, but should consent to as long as the system is trustworthy and fair to all. It's currently unknown whether some amount of secrecy will be necessary for effective defensive systems. Complete openness in offensive capabilities may not be acceptable to everyone no matter how beneficial it would be, and open publication of new weapon concepts may not always be desirable. Solving such problems and making such compromises requires further study.
	As long as all MNT capability is administered by an international body, with product designs being reported and tracked, it will be possible to verify the offensive and defensive capabilities of each nation. This approach depends on individual nations not developing independent MNT capabilities. Some designs will need to be kept secret. A small, diverse, trustworthy, collectively disinterested board of technology evaluators could assess the capabilities of each secret design: lethality, destructiveness, size, etc. The manufacture of each design could be tracked, allowing approximate knowledge of capabilities and intentions to be published without giving away secret details.
	If numerous countries do develop independent capabilities, it is hard to see how any country could be secure; even a massive (and wasteful) deployment of defensive MNT would not guarantee protection against new and unexpected kinds of weapons. As discussed in this essay by Thomas McCarthy, MNT's effect of cutting economic ties between countries greatly reduces their economic incentive to avoid war. And as discussed here by Mark Gubrud, and on CRN's Dangers page, an arms race interrupted by a preemptive strike is likely. Faced with such a scenario, all countries should be willing to accept mutual inspections to verify that there is no independent MNT development. However, the United States recently refused to allow biological inspections. Even if an international approach is the best hope for international security, it may be difficult to get it accepted. It is questionable, though, whether any one country such as the United States can successfully take on the role of international administrator.
Nanotech abundance is compatible with capitalist economy.	Much of society today is shaped by economy. There are at least three reasons to avoid disruption of the current economic model. First, a sudden change in economic activities would be quite destabilizing. Second, economic interests are quite powerful today and can probably prevent any plan they don't like. Third, capitalism is an excellent system for optimizing positive-sum problems, and the capitalist infrastructure is too useful to throw away. Opinions vary on whether networked file sharing (copyright violation) threatens the entertainment industry today, but the MPAA and RIAA are firmly convinced that it does, and have taken legal action (including lobbying) that has sometimes led to unexpected and unwarranted curtailment of freedoms. A distributed, low-cost manufacturing system could provoke a similar uproar. Embedded Security Management (ESM) can provide a platform for protection of intellectual property rights. In addition to security licensing, products could be restricted to be built only for customers who had paid for them. Since most of ESM is automated, this would not require a lot of resources. In addition, automatic scanning for designs that violate trademark, copyright, or patent would be useful to prevent illegal development of protected designs. (Copyright and patent law are very complex; sometimes separately-developed designs are OK, and sometimes they are not.) As explained below, this automatic scanning can also facilitate innovation.
	Consumers will want to purchase products and benefit from the major improvements that molecular manufacturing produces. Purchasing of designs will be quite easy—the main trouble will be finding the designs they want among the flood of new inventions. This implies that marketing and sales will still be necessary. Much of the economy in the United States today is based on service industries, and this need not change. Taking advantage of the nanofactory infrastructure, thousands of design companies may spring up, and may also develop from existing companies. With strong intellectual property protection, the cycle of innovation and purchase can continue, producing much profit for all involved and supporting a strong economy.
Capitalist nanotech is compatible with humanitarian relief.	There are billions of people in the world today who have almost no way of earning money. Many of these people are sick and even dying from malnutrition and disease, but may not be able to pay licensing fees for cheaply manufactured MNT products that would keep them alive. Global security, as well as humanitarian considerations, demand that their basic material needs be provided whether or not they can pay. There are many arguments that the owners of nanofactory technology should allow free use for humanitarian purposes. First, the profits to be made from selling water filters and mosquito netting are miniscule compared with the profits from selling high-end luxury goods. Second, if only one nanofactory design is allowed, this creates a monopoly, and monopolies can legitimately be regulated. Third, if billions of people can rapidly be raised from abject poverty, the global market for luxury goods will increase dramatically, which allows the owners to make more money (the "rising tide" argument). Fourth, both governments and charities should be willing to compensate the nanofactory owners handsomely for a blanket humanitarian license. Fifth, innovative products generate more money for the nanofactory owners—and to spur innovation, basic technologies should be free anyway. Sixth, if the future owners are not willing to agree to this at the time nanofactories are developed, they may be locked out of the development project in favor of those who will allow free humanitarian (and perhaps government) use. Seventh, lifesaving technology will be so cheap to produce that to restrict its use would be obscene; few individual business owners or stockholders would actually choose to prevent lifesaving use if they were directly confronted with the choice.
Innovative products, and control of new products, are both possible.	Even "mostly safe" products can be revolutionary by today's standards. The ability to pack a supercomputer into a sand grain allows all sorts of innovation. That combined with more capable sensors, displays, and actuators will allow amazing robotics to be developed. The range of products will be limited far more by human imagination than by technological restrictions. More risky products could be developed under careful scrutiny by licensed developers. CRN's ESM system allows a single nanofactory design to be used for both safe and risky products. Approval for production of especially risky products would be given only under carefully controlled circumstances. However, most product functions could be fulfilled within the safer categories. As nanofactory technology improves, new versions would have to be carefully checked to prevent technology leaks. However, even the first nanofactory will be able to build products with perhaps 50% of their theoretical maximum capability, so slow nanofactory improvement will not be a severe limitation.
	Decisions about which products to approve can be made at several levels. As noted above, MNT security monitors would be able to veto any product or class of products in order to prevent technology leak. Governments would be able to track the design and manufacture of products within their jurisdiction. Illegal product designs could be vetoed from further manufacture, and their designers arrested or blacklisted. Depending on the circumstances, people causing the creation of illegal products from foreign designs might also be detected and stopped or punished. There are many opportunities for a state to maintain control—a bigger problem seems to be avoiding the creation of a degree of control that violates human rights or allows government oppression such as blackmail and selective denial of service.
Environmental controls can be imposed, and remediation implemented.	Some products may not cause problems individually, but may cause problems when many copies are used. Environmental damage may occur from a variety of mechanisms, including construction of large buildings, deployment of large numbers of solar cells, release of heat produced by the operation of many nanodevices at high power density, and release of small particles creating litter that is hard to clean up and may be toxic at high concentrations. Such problems may not deter an individual user, so must be regulated collectively. A product (or type of product) that was sufficiently popular to be collectively damaging could be regulated through ESM, allowing only a certain number per person or per land area to be built.
	Molecular manufacturing will offer unprecedented mechanisms for alleviating existing air and water pollution and for cleaning up toxic waste sites. Working at the molecular level, such remediation will be far safer, more effective, and less expensive than today's techniques.
Patents, and patent reform, can be supported.	Innovation is important for several reasons. First, innovation will increase wealth and well-being as more useful products are created. Second, innovation will be a major driving force in the post-MNT economy. Third, people who are driven to innovate need an outlet for their talents. Nanofactory manufacturing provides an incredibly rich field for innovation. The basic unit of design is the nanoblock, less than a micron on a side; a cubic millimeter contains billions of nanoblocks, allowing an almost unlimited number of products. Maximum support of innovation depends on two factors: first, the field must be open, and second, specific inventions must be protected. This means that protection should not be applied in such a way that large areas of the field are unavailable. (Software patents have not had a good record in this regard.) Since a nanotech product design is no more or less than a specification of combinations of nanoblocks, automated design analysis systems can be an integral part of the patent system for these designs. This would allow immediate detection of patent infringers and of prior art, ensuring that patents are used as far as possible but not overused. A "patent holiday" of a few months would allow a base of prior art to be developed, ensuring that the most obvious and useful designs could not be owned by opportunists. However, a few months of invention will not even scratch the surface of the possible products; most inventions would remain to be developed and patented after the patent holiday expired.
Nanofactories can run off-grid.	Nanofactories are incredibly useful as a way of deploying advanced nanotechnology. A nanofactory would be self-contained, and would not rely on any 20th century infrastructure. Sunlight could be used for power—a solar electric generator can be built using only a few grams of diamondoid material per square meter. Feedstock material is currently unspecified, but lab-on-a-chip technology would probably allow locally available organic material to be processed into feedstock. (There isn't much carbon dioxide in the air; using that as a source of carbon would require processing huge volumes of air.) There is no reason why literally every person on earth should not have access to a nanofactory and its products. Humanitarian necessities could be free to all who needed them; luxuries could be far more luxurious, and far more lucrative for the owners and inventors of the technology, than today's crude products.
Almost all groups would have strong incentive to support this system.	Once the nanofactory is invented and deployed, and products are invented and made available, there will be little legitimate need for a competing technology. Every group that is benefiting from the nanofactory system stands to lose some or all of those benefits if a second system arises. (Consumers might enjoy lower prices, but would also suffer from lower security.) The potential damage to personal, national, and international security by a successful competing MNT program should inspire almost universal agreement that such a thing should not be allowed. Legitimate commercial use would be impossible, so no commercial entity would try it. Governments might want a covert nanofactory, but would not want any other government to have one; this should be sufficient incentive to submit to mutual inspections. Criminals would have many uses for powerful, untraceable products, and some criminal organizations have sufficient resources to finance a nanofactory program, especially since the difficulty will drop sharply once the first program is successful, and continue to decrease with time. Rogue political entities may have similar motivations and resources. Some amount of control of technology in general will be necessary to prevent the criminal development of independent nanofactories. However, a decade from now, this would merely be an extension of controls already in place to prevent terrorist development of weapons of mass destruction.
This proposal lays a flexible foundation for evolving administrative policy.	As MNT products and capabilities are better understood, and as defensive technologies (and possibly surveillance and monitoring technologies) are developed and deployed, it will be clearer how much control is necessary to prevent irresponsible use of molecular manufacturing. As discussed on our Technical Restrictions page, miniaturized MNT products allow the creation of intensive, semi-automated surveillance systems, which in theory can be implemented so as to preserve privacy. The degree to which such technologies are necessary will be determined by a much improved understanding of the destructive and defensive capabilities of nanotech products. Abusive use of such surveillance capabilities is quite possible, and we must hope that democracy and accountability will be able to prevent this. MNT will increase the ease of such surveillance but does not create the problem; even non-nanotech computers and surveillance devices will be quite powerful and may be ubiquitous within a decade. CRN cannot foresee what will be possible, or necessary, more than a few years after the invention of the nanofactory—the technology will become extremely powerful too quickly to forecast its effects on society, or society's effects on the direction of technology development. However, we believe that the policies outlined here will provide a means of surviving the first few years, and a sufficiently flexible foundation for whatever changes are needed in the subsequent years. Increased technological capabilities could reduce economic, environmental, and possibly political pressure, switching the emphasis of many of today's issues from allocating resources to maximizing wealth. In the same way, increased monitoring capabilities will decrease the need for intrusive and abusive police actions—even as they increase the possibility of extreme abuse. The kind of system we end up with will be the result of our choices. Almost anything is possible, and the great power of the technology demands a high degree of responsibility.

DEVIL'S ADVOCATE —

What if some group doesn't want to participate in a coordinated program of MNT development?

If they want to reject molecular manufacturing altogether, they should be allowed to—except where that would cause human rights violation such as unnecessary starvation. If they want to develop their own MNT, we don't think that's very safe. If most groups agree with our facts and conclusions, they will probably work to prevent independent development.

So you're proposing a ruling class that dispenses MNT to everyone else? Human nature guarantees that this will be abused.

Without our proposal, there are two likely bad possibilities. One is that MNT is unrestricted: everyone manages molecular manufacturing by themselves. That just looks too risky, especially in the first few years when we don't fully understand what MNT can do or how to defend against misuses of it. The other is that whoever gets molecular manufacturing first tries to set themselves up as world rulers; this is worse than what we're proposing. At least if we design it in advance, we can build in checks and balances between diverse interests, and try to avoid a single "ruling class".

Next Page: The Need for Early Development

Previous Page: The Need for Immediate Action

Title Page: Overview of Current Findings