Center for Responsible Nanotechnology
March 2004
Introduction
The mission of the
Center for Responsible Nanotechnology (CRN) is to raise
awareness of the issues presented by advanced
nanotechnology: the benefits and dangers, and the
possibilities for responsible use. We explore the ethical,
legal, and social implications (ELSI) of nanotechnology, and
its potentially disruptive consequence, exponential
general-purpose molecular manufacturing. The purpose of CRN
is to educate those who will influence the use of
nanotechnology, or be affected by it.
An important aspect
of this educational process is to create theoretical,
pictorial and textual representations of what may become
possible through nanoscale science and engineering (NSE),
especially though molecular manufacturing. CRN studies,
clarifies, and researches the issues involvedpolitical,
economic, military, humanitarian, and technologicaland
presents the results for both technical and popular
audiences, working to supply the information as effectively
as possible.
Our intention is to provide well-grounded and complete
information, clear explanation, and workable proposals that
support our vision of a world in which nanotechnology is
widely used for productive and beneficial purposes, and
where malicious uses are limited by effective administration
of the technology.
This effort
confronts numerous problems. Highly technical concepts must
be explained to a non-technical audience, often in a
text-only format. Our chosen topic of molecular
manufacturing suffers from premature hype, premature
debunking, and linguistic confusion resulting from the
various meanings of the word nanotechnology. We struggle
to maintain not only accuracy and simplicity, but also
credibility and clarity. Meanwhile, we deal with uncertainty
in scientific theory as well as in predictions of future
actions and choices. Yet when the stakes may be so high, the
effort is not only worthwhile, it is essential.
This paper reviews
the process CRN follows in choosing how and what to describe
as the likely results of our research into
molecular manufacturing. How do we select language that
will be accurate, informative, and compelling, while
promoting sensible and reasoned discussion? Is our strongest
motivation to perform technical work and present unbiased
results, or is it to develop and advocate for policy
positions that support our mission? Can it be both?
As the lines
between basic research, ethical responsibility, and advocacy
are blurred and debated, it is important for anyone involved
with NSE to ask themselves these questions. By examining and
describing our own experience, we hope to shed some light on
the subject.
Perspective
The first thing we,
and perhaps anyone in a position similar to ours, must do is
determine the perspective from which we will study a given
field and report on our findings. Are we to be dispassionate
observers, or concerned activists? The question is simple to
pose but complicated to answer. It runs to the very nature
of study and observation.
To begin with, can
any observer stand fully outside of a subject field, or does
the act of observation constitute an unavoidable
involvement? We know that on the level of quantum physics,
observation not only gathers information but also imparts it
to the observed subject. Does this fundamental fact apply on
a metaphorical basis to observers at the macro level?
But beyond these phenomenological questions, we must ask:
What motivates us to engage in our analysis of
nanotechnology and its societal implications? What
preconceptions about these issues do we bring with us to the
act of observation? What principles will guide us in
analyzing and reporting our findings?
As we at CRN
examine our motivations and practices (which is something we
do on a regular basis), frequently we find that our
preferred position is to be in the middle between opposing
extremes. In this case, that means we do not view or present
ourselves as completely dispassionate observers (if such a
thing is even possible), nor do we desire to behave as or to
be perceived as zealous activists. Rather, we hope to act
and be accepted as informed, principled, interested analysts
and, ultimately, effective advocates.
We approach the
analysis of advanced nanotechnology armed with knowledge
acquired through years of both specific and broad studies,
of both a technical and general nature. Our current
understanding has led us to certain conclusions that serve
as motivation, but our principles also demand that we
examine and consider arguments that may contradict our
current understanding.
The major
motivation behind our work at CRN is the conclusion that
molecular manufacturing almost certainly will become a
reality (through one of several technological avenues)
within the next twenty years, and perhaps much sooner. Our
studies suggest that this new general-purpose technology
will have a significant impact across nearly all segments of
society. It likely will be transformative, and could be
highly disruptive.
If the technology
does prove to be as powerful as we expect, its easy to see
that effective use of molecular manufacturing could be
widely beneficial, and that unwise use could be highly
dangerous. The extreme degree of the technologys potential
will tempt extreme reactions and already is resulting in
extreme proposed solutions. We are motivated, therefore, to
research, propose, and advocate for policy positions that
will allow for safe use of the technology while avoiding the
societal damagethe human costof reactionary solutions.
This approach often finds us pitted between those opposed to
regulation in any form and those arguing for what we would
consider to be unsupportably onerous restrictions.
Our conclusions
about the potentially transformative and disruptive nature
of molecular manufacturing are by no means widely accepted.
It is far easier to find disagreementsometimes
vehementfrom influential persons in government, business,
and academia, than to find sympathy with our positions. Does
this mean that we are wrong? Obviously, it does not,
although of course it also does not mean that our
conclusions are necessarily correct. The annals of history
are replete with figures who struggled against the
establishment until their iconoclastic ideas were finally
proved correct, often posthumously. But there are doubtless
many more persons lost to history whose unpopular ideas
proved to be fallacious. So the unpopularity of our ideas
does not signify anything about their correctness.
However, the fact
that so many learned people are convinced we are wrong
should lead us to carefully consider our positions and
examine them for error.
Principles
Three particular
principles are required for an effective examination of our
positions. They are: 1) a dedication to the free exchange of
information; 2) a desire for constructive dialogue with
critics; and 3) a willingness to be wrong.
Of the three major
types of organizations, Guardian, Commercial, and
Information (Phoenix
and Treder, 2003), CRN definitely is an
Information-ethic organization. Our function is to produce
information and publish it widely. Unlike Guardian
institutions, we will attempt to be open about
everythinghence a self-examining and revealing paper such
as this oneunless there is an overwhelming reason to keep
something secret. Unlike Commercial institutions,
Information organizations are not focused on money; we view
money as simply a means to an end. Our motivating principles
include building a solid reputation, being known according
to our work, and being distinguished by our unique
contributions.
CRN operates on the
belief that an understanding of future technical
possibilities will be vital in order to prepare for smooth
adoption and responsible use of new technologies, and to
allocate research attention and funding appropriately.
Estimates of nanotechnology's ultimate potential, and the
timeline and cost for development, vary widely, to say the
least. But information is power; only through intensive
studies can we ensure that the developers and the future
administrators of this powerful capability have the tools
they need to make the right decisions. A detailed
understanding of molecular manufacturing technology is
necessary to prepare for its eventual development.
So we are dedicated
to open exchange of information, we are motivated by the
need for solid research to assist in the decision-making
process, and we seek to understand opinions that differ from
ours. We will admit when we are wrong and gladly will change
our positions to something more clearly correct when that is
indicated.
Presentation
With all this as
background, we can address the specific issues of
terminology and descriptive language chosen for use by the
Center for Responsible Nanotechnology. In a recent email,
CRN co-founder Chris Phoenix lamented:
Perhaps we need
to work on our communication skills. Whenever we propose
anything, it seems like people hear it as suggesting the
extreme, although that's usually not what we mean. A
major meta-strategy of CRN is to be middle-of-the-road
on almost everything, recognizing that extremes are very
likely to be a bad idea for one reason or another. But
somehow that message does not always come through
clearly.
The challenge of
selecting language involves not just technical accuracy but
also effectiveness in communicating underlying ideas and
intent. As noted above in the section on Perspective, CRN
aims to be something between dispassionate observers and
zealous activists. Similarly, were trying to carve out a
position between being boosters for nanotechnology,
pushing for progress at all costs, and being sentinels,
raising awareness of potential dangers. We recognize the
great promise of the technology to relieve human suffering
and create unprecedented abundance, and we'll do whatever we
can to bring that about. At the same time, it would be
irresponsible of us not to study the inherent risks, report
our findings, and suggest solutions.
Our internal debate
over how to describe the field in which we work is
instructive. We are called the Center for Responsible
Nanotechnologynot molecular nanotechnology, or advanced
nanotechnology, but simply nanotechnology. Is this an
appropriate name, or is it misleading?
The word
nanotechnology has not yet acquired a common meaning.
Widely disparate definitions can be found in dictionaries,
organizational glossaries, and published documents on the
Internet and in print. Usage of the term in science fiction,
both credible and fantastic, along with the sometimes
questionable adoption of the word by research facilities and
companies seeking funding or investment contributes to this
semantic dilution.
Because the main
focus of CRN is on the results of a type of nanotechnology
that may not exist for another decade or two, there could be
some confusion if people think we are working to promote
responsible use of present-day nanoscale sciences and
products, such as paints, fabrics, coatings, or rocket
fuels. Should we therefore have called ourselves the Center
for Responsible Future Nanotechnology, Advanced
Nanotechnology, or Molecular Nanotechnology?
A short name
obviously is preferable over a long one. It also seems clear
that to the large majority of people, nanotechnology means
something more exciting and futuristic than stain-resistant
pants. It meansit is synonymous withwhat we might more
precisely call advanced nanotechnology or molecular
nanotechnology.
In CRNs early
writings, during the first half of 2003, we made frequent
use of the term molecular nanotechnology, and the
abbreviation MNT. This was intended to distinguish our
longer-term expectations for the field from the broader
application of nanotechnology that had become current as
government funding was made available and applicants started
labeling even mundane types of research as nanotechnology.
But around the
middle of 2003, it became apparent to us that MNT and
molecular nanotechnology possessed a negative connotation
with many serious researchers. The terms were associated
almost invariably with fantastic notions like bloodstream
nanobots, true universal assemblers (meat machines), and
theoretically ubiquitous utility fog. Such concepts
admittedly are fascinating to consider and someday may
become reality, but they seem to be further in the future
than are the middle-period developments that concern CRN.
Much of this
connotative difficulty can be traced to the order in which
Eric Drexler (who properly can be called the father of
nanotechnology) introduced his concepts to the scientific
community and the world at large. His first published book
was Engines of Creation (Drexler
1986), in which he laid out the spectacular
possibilities of this anticipated future technology,
including universal assemblers that would let us build
almost anything that the laws of nature allow to exist.
This promised the end of hunger, fine control of nature,
mastery over space, and even glimpses of human immortality.
Engines was a popular success and can be credited with
inspiring many of todays nanoscale scientists, including
Richard Smalley, to enter the field.
It wasnt until six
years later that Drexler published Nanosystems: Molecular
Machinery, Manufacturing, and Computation (Drexler
1992), a far more rigorous and detailed analysis of the
science and technology that would be required to turn some
of these far-out concepts into near-term reality. In
Nanosystems, the focus was on the early stages of
nanotechnology manufacturing and the tone was more sober and
scholarly. Nevertheless, the die had already been cast, and
Drexler hereafter was labeled by many in the scientific
establishment as a visionary dreamer, and not someone to be
taken seriously.
There are numerous
critics of Drexlers ideas who only have read (or read
about) Engines of Creation and never have studied
Nanosystems. One wonders how things might be different
today if the books had been published in the opposite order.
The words and
phrases we use clearly will communicate more than just their
particular meaning. Context and connotation also must be
considered. As an example, every time we use the name
Drexler in a document, it has an effect on readers beyond
simply being the name of a person. Depending on their
awareness of the man and his work, and their opinion about
it, inclusion of this name can in itself communicate
significant meaning. The reader instantly may become
favorably or unfavorably disposed toward CRN just by seeing
the name, especially if it is used in context of which the
reader strongly approves or disapproves.
Definitions
CRNs research is
concentrated on what might be called the middle period of
nanotechnology development, the point between todays
non-manufacturing NSE technologies and the sci-fi visions
of Engines of Creation (note again the recurring
theme of being in the middle).
In response to the
negative associations of molecular nanotechnology and
MNT with visionary universal assemblers, we made an
attempt in the latter half of 2003 to distinguish this
middle period as dealing with a limited version of molecular
nanotechnology, or LMNT. This was characterized as
implementing just a tiny fraction of possible chemistry,
aimed at achieving a limited molecular manufacturing
capability based only on carbon lattice
configurationsdiamond, graphite, and fullerenesknown
collectively as "diamondoid". We found, however, that
although this is a useful and important distinction for
technical writing, the meaning is too arcane and derivative
for general usage.
Near the end of
2003, CRN decided to replace most usages of molecular
nanotechnology in our writing with molecular
manufacturing. This was thought wise not only to avoid the
baggage associated with MNT but also to more specifically
identify the period when large-scale manufacturing of
products at the molecular level has become possible. To be
more precise and descriptive, we sometimes will use the
fuller phrase exponential general-purpose molecular
manufacturing. Exponential refers to the capability of the
technology to reproduce its own means of manufacturing
(self-copying). General-purpose suggests that the technology
has application across a broad spectrum of industries and
hence will affect many segments of society.
Recently announced
developments in nucleic acid engineering make it clear that
our choice of the term molecular manufacturing is a good
one. CRN now defines molecular manufacturing as any
technology that implements digital operations, nanoscale
construction, self-manufacture, programmable properties, and
low error rates, and this definition can apply to any
technologydiamondoid or notthat meets all five criteria (Phoenix
2004).
Digital
operations means that each
manufacturing process has a well-defined discontinuity
between success and error. If a certain design is
constructed multiple times, the products that do not contain
definite errors will be identical. This implies high
reliability and predictability for the error-free copies.
Nanoscale
construction means that the
chemical building blocks can form, either singly or in
combination, features in the 1-100 nanometer size range.
Since no molecule is perfectly stiff, the physical
arrangement of the features will not be perfectly precise.
The permissible degree of uncertainty will depend on the
application, but at least some physical coherence will be
necessary for self-manufacture.
Self-manufacture
means that the chemical system's range of designs must
include devices that can contribute to the manufacture of
other designs in the range. The functionality may range from
flexible templating to nano-robotics doing pick-and-place
operations. Self-manufacture may significantly lower the
cost and increase the complexity of products, especially if
it can be automatedwhich is made easier by digital
operations and low error rates.
Programmable
properties means that low-level
designs can be specified or computed by describing
higher-level features. Within a certain range, the design
space will accommodate any specified feature without
additional research. Essentially, this means that design
rules and levels of abstraction can be used in the design
process. A wide variety of features can be successfully
specified without chemical research.
Low error rates
means that the manufacturing process, and the subsequent
operation of the products, has a usefully high success rate.
Error rates may vary by many orders of magnitude. For
example, a rate of 10-12 would be very poor for digital
transistor logic, but a rate of 10-3 would be excellent for
organic chemical synthesis. In general, an error rate per
operation (e.g. per atom added to a product) of 10-9 to
10-12 may be adequate, though better rates may be achievable
(Merkle
1997).
This new definition
of molecular manufacturing is important and timely because
nucleic acid engineering appears to be moving rapidly toward
satisfying the five specified criteria (Shih
et al 2004). It probably will be a few orders of
magnitude less powerful than diamondoid, but still many
orders ahead of today's manufacturing technologies. Until
now, we have only had one technologyoriginal Drexler-style
MNTto evaluate. But with two possible molecular
manufacturing systems to compare, it's easier to talk about
the performance tradeoffs of each technology.
Although no
technology today qualifies as molecular manufacturing, each
of the specified requirements is implemented in some
currently existing technology, and at least two NSE
technologies are developing rapidly toward a convergence of
all five criteria. We now have a definitional framework with
which to judge these and other new approaches that may be
developed.
Conclusion
This paper has
explained CRNs ongoing process of defining and describing
our work. By carefully and repeatedly examining our
terminology, we hope to succeed in walking the narrow middle
line between dispassionate observation and zealous activism;
between being boosters for nanotechnology and being
sentinels. We aim to avoid being marginalized as irrelevant
fanatics, and instead fulfill our chosen function as
informed, principled, interested analysts and effective
advocates for responsible use of nanotechnology.
REFERENCES
Drexler, K. Eric (1986) Engines of Creation (Anchor Books).
Drexler, K. Eric (1992) Nanosystems: Molecular Machinery,
Manufacturing, and Computation (John Wiley & Sons).
Merkle, Ralph C. (1997) A New Family of Six Degree of
Freedom Positional Devices, Nanotechnology, 8(2), 47-52.
Phoenix, Chris and Treder, Mike (2003) Three
Systems of Action: A Proposed Application for Effective
Administration of Molecular Nanotechnology
Phoenix, Chris (2004) Studying Molecular Manufacturing, to
be published by the Institute of Electrical and Electronics
Engineers.
Shih, William M., Quispe, Joel D., and Joyce, Gerald F.
(2004) A 1.7-kilobase single-stranded DNA that folds into a
nanoscale octahedron, Nature.