Responsible Nanotechnology

Is matter digital?

Richard Jones recently wrote a blog post ending with the statement, "Matter is not digital."

Jones was responding to Eric Drexler's statement to the NAS that, just as computers are amazingly useful general-purpose computation machines (e.g. capable of displaying movies), molecular manufacturing could lead to general-purpose manufacturing systems making a very large range of products. Jones argued that, "For the idea of general purpose manufacturing to be convincing, one would need to believe that there was an analogous way in which all material things could be represented by a simple low level code. I think this leads to an insoluble dilemma - the need to find simple low level operations drives one to use a minimum number - preferably one - basic mechanosynthesis step."

A much more detailed analogy can be drawn between molecular manufacturing (MM) and computers. And it shows that the issue is not whether matter "is" digital, but whether digital design principles can be applied to matter. After all, movies are not digital. But they can be handled digitally.

Transistors, the basis of modern computers, are not digital. They are nonlinear, meaning that it takes very little effort to make them switch from mostly-on to mostly-off. That lets them ignore the error in the signals (in exchange for some energy), allowing us to design circuits as though the signals really were a digital ideal.

Forces between atoms as they bond are also nonlinear. As you push them together, they "snap" into position. That allows maintenance of mechanical precision: it's not hard, in theory, for a molecular manufacturing system to make a product fully as precise as itself. So covalent bonds between atoms are analogous to transistors. Individual bonds correspond to the ones and zeros level.

Computers are not designed in terms of individual transistors. They use microprocessors, which contain thousands to billions of transistors. The microprocessors implement a few dozen instructions called "assembly language" which are carried out repetitively. This is analogous to the mechanosynthetic reactions. You don't need as few as possible. You need a small but convenient set. A system which could carry out a few dozen synthetic reactions in programmable sequence could be very useful.

Jones continues: "But in limiting ourselves in this way, we make life very difficult for ourselves in trying to achieve the broad range of functions and actions that we are going to want these artefacts for. Material properties are multidimensional, and it’s difficult to believe that one material can meet all our needs."

By distinguishing between the nonlinear, precision-preserving level (transistors and bonding) and the level of programmable operations (assembly language and mechanosynthetic operations), it should be clear that the digital approach to mechanosynthesis is not a limitation, and in particular does not limit us to one material. But for convenience, an efficient system will probably produce only a few materials.

Now let's consider the separate question of whether a few programmable material-building operations are suitable for producing a wide range of output types.

For nanoscale operations like binding arbitrary molecules, it remains to be seen how difficult it will be to achieve near-universal competence. But most products that we use today do not involve engineered nanoscale operations. Instead, they work at a scale of many microns. If we can design and build products containing machine systems far smaller than a cell, then we should be able to make virtual materials with properties that can be engineered over a wide range.

For example, a parameterized nanoscale truss design could produce structures which on larger scales had a vast range of strength, elasticity, and energy dissipation. A nanoscale digital switch could be used to build any circuit, and when combined with an actuator and a power source, could emulate a wide range of deformable structures. A few designs for photon handling, sensing (much of which can be implemented with mechanics), and so on should be enough to build almost any reasonable macro-scale product we can design.

Of course, the design of complicated products won't be simple. No one writes a modern user interface in "down to the bare metal" assembly language, but instead uses object orientation, scripting languages, an operating system, and a ton and a half of libraries. A product like a car will likewise not be programmed "down to the bare atoms." But assembly language is not a limitation but a foundation; it is general purpose, and all the tools of software are built on top of it. Likewise, the mechanosynthetic reactions and the nanosystems they build will not be a limitation, but will provide a foundation for endless recombination, elaboration, and hierarchical design.

Yes, matter can be treated digitally: bonds make molecules, molecules make machines, machines can be engineered. This is the strength of molecular manufacturing.

Chris

General-Purpose Manufacturing Scorecard

With all the nanoscale research going on today, in many diverse fields, how can we tell which of it is relevant to molecular manufacturing?

I've written a draft of a scorecard to determine how relevant a piece of research is for general-purpose manufacturing. The draft is published on Wise-Nano.org.

The scorecard does not actually refer to nanotechnology or molecular manufacturing -- I tried to make it as widely applicable as possible. You could rate a semiconductor fab or a plastic-squirting rapid prototyping machine with this scorecard -- and they would rate higher than a lot of today's nanoscale manufacturing research. My intent is that this scorecard will be general enough to be "non-partisan," but will clearly distinguish approaches that are useful for molecular manufacturing (MM) from those that are special-purpose, unsuitable, or irrelevant.

In theory, a non-MM technology could get a high score. If such a technology is identified, then either I need to fine-tune the scorecard, or MM is not as unique as I think.

As I write this, I haven't yet figured out a way to formulate a question that addresses the issue of system and product cost. Please chip in and help me improve it.

(Please comment/contribute at the Wise-Nano page. For those who would rather talk here: Unless you request otherwise, posting a comment to this thread constitutes releasing it under a Gnu FDL license, so I can copy it to Wise-Nano.)

Chris

Challenges and Opportunities for Developing Countries

As reported before, last week I attended an Expert Group Meeting titled "North-South Dialogue on Nanotechnology: Challenges and Opportunities." This meeting was organized by the International Centre for Science and High Technology (ICS), an Institute of the United Nations operating in the framework of the United Nations Industrial Development Organization.

The organizers' introductory background description says:

Although nanotechnology is in its infancy, now may be the right time to explore its implementation in developing countries (DCs). Most of the current debate in industrialized countries is focused primarily on nanotechnology risks rather than its potential benefits. While it is true that there are legitimate risks that need to be addressed, there is also a need to adequately consider and understand how nanotechnology can bring benefits to knowledge and the economy in DCs. Moreover, an exclusive focus on the risks will create and widen another gap, similar to the digital and biotechnological gaps, between industrialized countries and DCs.

A total of eighty people from all around the world — scientists, academics, government personnel, and NGO representatives — gathered in Trieste, Italy, for this event. We had two full days of presentations and group discussions, followed by group dinners at which many more ideas were exchanged. On the third day, we broke into four smaller groups to develop recommendations on how the ICS might implement some of the proposals put forward during the meeting.

To help us prepare, the ICS published an Aide-Mémoire, which stated in part:

…in this moment there is no global effort to promote awareness of the potential applications of nanotechnology for development, to create new regulatory regimes (or build upon existing ones) for managing nanotechnology's associated risks and for promoting global public goods, to provide a platform for constructive dialogue among all stakeholders including representatives from government, industry, academia and local populations.

I used this theme for the presentation I delivered, which has now been posted on our website. My emphasis, of course, was on the possible need for administrative structures to manage advanced nanotechnology, and some of the factors and methods that might be considered in creating these "new regulatory regimes."

Here are a few of the many interesting people I got to know during our three days there:

- Roop Mahajan, professor of Mechanical Engineering at the University of Colorado
- Todd Barker, one of the founding partners of Meridian Institute
- Uzi Landman, a computer modeling specialist with the Georgia Institute of Technology
- Tom Kalil, former science advisor to President Bill Clinton
- Chris Bunting, General Secretary of the International Risk Governance Council
- Terry Turney, Science Director at CSIRO Manufacturing in Australia
- Angela Hullman, European Commission Project Officer for Nanosciences and Nanotechnologies
- Ray Pimentel, Vice Consul with the British Consulate in Chicago
- Mark Morrison, Nanoforum Coordinator for the UK's Institute of Nanotechnology
- Don Maclurcan, with Australia’s Institute of Nanoscale Technology
- Kathy Jo Wetter, an ETC Group researcher
- Catherine Brahic, a staff journalist for SciDev.net

Three attendees were people I already knew:

- Mihail Roco, head of the U.S. National Nanotechnology Initiative (NNI)
- Sinclair Wang, head of Taiwan Nanotechnology, and a member of CRN’s Board of Advisors
- Judith Light Feather, head of The NanoTechnology Group Inc.

I was gratified to discover how well known CRN already is within the worldwide nanotech community. Many people came up to me and said they knew about us, and many reported that they regularly read this blog.

Overall, it was a rewarding and highly constructive event, brilliantly planned and carried out by Dr. Luisa Mestroni and her hard-working, talented staff at the ICS. More gatherings of this sort will be imperative as we come closer to the time when exponential general-purpose molecular manufacturing becomes a reality.

Mike Treder

By Any Other Name

A rose by any other name would smell as sweet -- William Shakespeare ("Romeo and Juliet")

Recent discussions -- on our blog, on Howard Lovy's Nanobot, and on the Soft Machines blog -- about shifting perspectives toward the possible development of molecular manufacturing reminded me of something I'd read last April.

Adrian Tymes, who attended an NNI conference in Washington DC, reported that they actually seemed to be aiming toward what we define as molecular manufacturing, even if they did not call it by that name. Rather than repeat myself, I'll suggest that you read his report, and our commentary on it.

Ten years away?

Mike Treder

Shifting Perspectives

In my discussions with people and in recent presentations, I've been saying that when we look back, 2005 will be seen as a pivotal year in the development of advanced nanotechnology -- the capability for precise, general-purpose, exponential manufacturing. This expectation is based on conversations with influential leaders who acknowledge that high-level views toward this radical technology are gradually changing from outright dismissal to cautious consideration.

At the ICS/UNIDO Expert Group Meeting on nanotechnology in which I participated (and will provide a fuller report in a day or two), several people pointed out to me that what they are now hearing said is not that molecular manufacturing is impossible, but that it is at least 30 to 50 years away. That's a big shift in perspective.

More of this new outlook can be seen in the superb reporting provided by Adam Keiper on the National Research Council committee hearings in Washington (which Chris Phoenix attended and addressed)...

During the workshop’s final minutes, Eric Drexler made closing remarks arguing that "at this point, there is a large weight of evidence -- detailed quantitative analysis -- indicating that large-scale molecular manufacturing is feasible." More noteworthy than Drexler's remark was the response to it from Clayton Teague, head of the National Nanotechnology Coordination Office: "I agree with you very much, and I think you'll find everyone on the NNI thinks that it is something of international importance."

Most of the day was characterized by similar attitudes of agreement and agreeableness, with a few supporters of molecular manufacturing expressing their considerable surprise at the committee’s "open-mindedness." Although most of the committee's work is done behind closed doors and no tapes or transcripts of this workshop will be made available to the public, the plain fact is that no recording could capture the full atmosphere of the three-day workshop, in which some of the most important work happened off-the-record, in the conversations over meals and during breaks between panels. There were indications of conciliation and comity, and there seemed to be new avenues of communication and cooperation, between representatives of the government, nano-commerce, and the molecular manufacturing community.

This seems to be a promising start, and CRN looks forward to much more open dialogue about the needed technical research and policy issues that should be addressed.

Mike Treder

Nanotechnology Questions

Recently we have received several questions here from people interested in learning more about nanotechnology and career prospects, or about general nanotech progress in broad fields such as computers or medicine.

We'd like to be able to respond, but as stated in our message about mechanical engineering, our area of specialty is molecular manufacturing and the policy issues raised by that transformative near-future capability.

Almost all of the research and development occurring today in nanotechnology -- whether related to mechanical engineering, chemistry, pharmaceuticals, materials science, computing, medicine, or other areas -- is not directly connected to CRN's main focus. Therefore, we can't comment authoritatively on most of that work, nor on career opportunities.

For general information about today's nanotech, you may want to begin your research at Nanotechnology Now; for employment opportunities, try tinytechjobs.

Research Related to Molecular Manufacturing

Molecular manufacturing (MM) is still in an early stage of development. Despite some lab demos of relevant capabilities, it's not even known yet whether the first useful system will be based on polymers formed in water or crystalline solids formed in vacuum. So can we say for sure whether any nanoscale technology is irrelevant to the development of MM?

We can at least say that some technologies are a lot less likely to be relevant than others. The goal and method of MM is atomically precise fabrication of molecular structures under direct mechanical computer control. Many nanoscale technologies have nothing to do with this and will not help to achieve it. That's not to say they're not useful or interesting; just that they're not related to this branch of nanotechnology.

Many manufacturing technologies are not atomically precise and don't seem likely to be made so. For example, bulk processes for producing nanoparticles can produce particles that are the right size within a fraction of a nanometer--but that is hundreds of atoms. Lithography may become atomically precise--in three decades or so.

Here are some questions to ask:

Blueprint delivery: How much information can be delivered to the nanoscale and incorporated in the product? (How many parameters can be changed, and how rapidly?)

Is the technique programmable--can you change parameters and get a predictable result?

Does the technology take advantage of the inherent precision of covalent bonds and molecular structures so as to produce precisely identical products?

Does the research help to develop molecular or at least nanoscale mechanics?

If manufacturing, is the technique scaleable to produce large amounts of product? If sensing, can it produce high bandwidth of information?

How small can the tools be miniaturized? To the nanoscale, or at least to micron-scale?

Does it promote the goal of tools that can contribute to the manufacture of duplicate tools?

The vast majority of research funded under the U.S. National Nanotechnology Initiative (NNI) does not score very high by these criteria. Again, it may be very useful for curing diseases, building faster computers, and so on. I am not trying to argue that any of it is not worth doing. But most of it is interesting for other reasons than its connection to molecular manufacturing. The NNI's research, with very few exceptions, is simply not targeted at that goal.

Chris

Molecular Manufacturing, Nanoscale Stuff, NNI

[Updated to remove personal references]

How do you tell if your billion-dollar research program will contribute to developing molecular manufacturing (MM)?

The broad goal of MM is to build lots of innovative stuff at the nanoscale and make it do what you want, right? And since a specific proposal for how to develop MM doesn't exist yet, almost anything to do with the nanoscale might turn out to be helpful, right?

Without some careful study of molecular manufacturing theory, there's no basis for more specific criteria. And someone who hasn't done that careful study might simply miss the fact that there's anything to study.

The National Nanotechnology Initiative (NNI) looks pretty bad to the MM community. Several of the NNI's most prominent spokesmen and leaders have said that MM is ridiculous, impossible. But they claim to have the same goals as MM, and be working toward those goals. Meanwhile, they fund lots of research that clearly has nothing to do with MM, and almost none that's helpful. Now they've started listing "nanomanufacturing" as one of their Program Component Areas. But we know they don't like MM--they actively sabotage it--so they must be trying to muddy the waters by pretending they're working on it when they're not. In fact, this accusation was made on the last day of the recent NAS meeting.

But--suppose some people in NNI leadership think that the NNI is doing a good job of working toward complete technical competence at the nanoscale. And suppose that they have never paid all that much attention to MM, except to note that the goal is the same. Suppose they don't realize that the MM style is rather different from most of the nanoscale technologies they are funding. Suppose they think that the rhetorical attacks of the past few years against MM are just squabbles over details of method rather than attempts to shut down an entire subfield. In short, suppose they've never learned enough about MM to realize that they should learn more about it.

Is this line of supposition self-consistent? It seems so. It requires assuming that the NNI does not have a unified understanding of MM: some people there think it's a narrow, politically dangerous, and technically implausible proposal that should be squashed, while others think it's a very broad goal that can best be achieved by broad research. It seems quite plausible that people involved with the NNI would be uncoordinated on some issues, maybe including MM. And MM is elegant once you learn it, but it's not intuitive, and there haven't been any tutorials available that would be suitable for a busy administrator. Several days after the conversation, I'm still thinking about how to explain how to judge which research is relevant--and I've been studying MM for more than a decade.

My current working hypothesis that almost no one on either side actually understands each other's position. The solution is communication.

Chris

Good News from Washington

I just got back from the National Academy of Sciences meeting that was mandated in the 2003 Nano Act to study "molecular self-assembly." As far as I can tell, it went extremely well. I spoke in the second session, on definitions, but I was there for all six sessions. More details on that later; for tonight I'll just give initial impressions.

The committee focused directly on molecular manufacturing--which is good, because the wording in the near-final version of the law was "molecular manufacturing" before a nano business organization lobbied to change it to scuttle the study.

(In what follows, MM is molecular manufacturing; NNI is National Nanotechnology Initiative.)

There were some skeptics on the committee, and some who seemed more willing to accept that MM could work. But everyone there was clearly very smart, and everyone seemed intellectually honest. I'm pretty confident that MM will get a fair hearing--that they won't write a report which merely endorses the unscientific and off-topic objections that have been floating around for the last decade or two. Whatever they produce will be the result of open-minded and very careful study.

At the end of the second day, the committee went around the table with each person giving their personal reactions to what they'd seen so far--two sessions of history, one of definitions, and two of technology. A few members had to leave early, and a few said they were still absorbing what they'd heard. I took notes, but not quotes, and I was scribbling in a hurry, so I won't give people's names lest I mis-attribute an idea to them. But this is what I heard:

A thousand miles starts with a single step. We're busy proving A->B. We have no proof for E->F. (Where F is molecular manufacturing.)

There was impressive research in the last session. It somewhat alleviated my fears [about MM being impossible?]. We need a wide net for nanoscale tools--we don't know where is the tipping point for a breakthrough. Maybe we need a DARPA-like high risk/high reward pot to see what's possible.

I did biopolymers and thought it was nanotechnology; now I see we need definitions.

I've been through tech revolutions--lasers and computer chips. Big dreams are good; hats off to Drexler. However--it drives me crazy when people do computations without taking physical issues into account. If you say "manufacturing" you inherit physical issues--tolerances, yield, etc.

I work with nanotech companies--I see and appreciate proofs of concept--there are lots of [research?] directions. We need to survey the landscape widely to avoid being surprised by competitive innovations.

We're hearing that MM is a done deal--but we're being asked to consider if it's possible. I can't forget big picture/system. The devil is in the details. Why don't we have advanced e.g. medical products already?

The books Discoverers and Creators talk about history of technology. We owe a debt of discovery / another chapter in those books to Drexler; he's why we're sitting here. We're at a critical point: what will that chapter look like?

Thanks to the professors [in the tech sessions] for showing that the first steps are done. I advise patience to the advance thinkers. I did a heretical research. In 1972 I was told: show the first step. I did it. Then the engineers didn't care. So I showed them samples. *Today* my material is first being used in jet engines. (I had given up in '96.) We need demonstrations and patience.

I'm a scientist and engineer. Want to thank the panelists and visionaries. And Roco, for vision. Nanomanufacturing is in the NNI vision. [Drexler interrupts: nanomanufacturing is different from MM--the NNI is trying to muddy the waters.] I hope we can find a way to get materials into jet engines more rapidly. [Previous speaker interrupts--No! Be conservative with jet engines!] Let's se a grand challenge: an artificial ribosome. How stripped-down can we get a cell with a reasonable error rate for manufacturing? Keep a comparable goal in mind.

We're still in the process of identifying tasks. I did structure/function relationships of proteins and enzymes. Cells do have error correction--I'm not sure we're near achieving that. We need much more terminology work.

The NNI has come a long way. There's a lot going on.

Thanks everyone. I appreciated the civil tone of the discourse, and the enthusiasm. The passion for ideas was good to see. As someone else said: we haven't talked about: is MM desirable? At some point we need to look at societal needs we could address with this tech [something about public funds] - given fundamental research, keep in mind "to what end?" To the extent researchers can do that, it'll be helpful - resonant - public issues: health, energy.

I see progress toward MM: biology, chemistry, mechanical. We need to know more about fundamental limitations; I was impressed by a presentation on thermal noise.

I'm sure different people will see different things in the meeting, and these brief personal summaries don't even begin to give a full picture of the structure and content of the meeting. But I think they show the spirit of questioning that we need to begin a full and careful discussion of the technical merits of molecular manufacturing. I don't expect the committee to say either "Yes it's workable" or "no it's not." But "We looked at it for months and couldn't rule it out" would be a significant step forward.

Chris

CRN in Italy

I'm writing this entry from the beautiful city of Trieste, in the northeast corner of Italy, where I am attending an Expert Group Meeting on "North-South dialog on nanotechnology: challenges and opportunities." The meeting is organized by the International Centre for Science and High Technology, an Institute of the United Nations operating in the framework of the United Nations Industrial Development Organization.

Yesterday afternoon I delivered an invited presentation on the possible need for new regulatory structures to manage advanced nanotechnology. I'll write more about my trip, the outcome of the event, and my reflections on the current views of nanotechnology when I get back, which will be in a few days.

Mike Treder