I, Nanobot
The coming elimination of the barrier between living and nonliving materials will lead to "animats" (living materials) -- nanobiotechnology devices that can survive and function inside human beings, derive energy from biological metabolism, and copy themselves by molecular self-assembly. When that moment happens, it very likely may be beyond our control....
Originally published on Salon.com
March 9, 2006. Reprinted with permission on KurzweilAI.net July
7, 2006.
Don't call me Ishmael, for I am not a survivor. Don't call me Cassandra
either, since some might believe what I foretell. Perhaps I am the
final manifestation of the singularity ignited in Olduvi
Gorge a million and a half years ago. The flame that has grown
to consume our planet and send sparks into outer space. The singularity
that started as an ineffable, ineluctable pulse resonating through
the neural matrix of Homo
habilis. A voice that said, You whoever you are, You must sharpen
that stone, pick up that bone, cross that line. A voice of supreme
paradox; one that simultaneously makes us uniquely human, yet is
itself not human. Nor is it the black extraterrestrial monolith
of Stanley Kubrick's imagining. Rather, it was always here. Hard-wired
into us at the atomic level—and we into it. A voice whose physical
manifestation, the tool, sang its song millions of years before
human beings walked the earth. This voice prophesied and then enabled
our coming. It will instruct us in our going. Or so I say, while
understanding too well that in the 21st century we are all jaded
and stultified with sensory overload. It's always the end of the
world as we know it—and we feel bored.
So why listen to the voice of one who is not Ishmael, not Cassandra,
not even Ralph Nader? Because I can tell you something that no one
else can. I can tell you the exact moment when Homo sapiens will
cease to exist. And I can tell you how the end will come. I can
show you the exact design of the device that will bring us down.
I can reveal the blueprint, provide the precise technical specifications.
Long before we can melt the polar ice caps, or denude the rain forests,
or colonize the moon, we will be gone. And we will not—definitely
will not—end with a bang or a whimper. The human race will
go to its extinction in a state of supreme exaltation, like an actor
climbing the stairs to accept an Academy Award. We will exit the
stage of existence thinking we are going to a spectacular party.
The usual suspects—those who have become known for predicting
the evolution of humans and their technology—just don't get
it. Mainly because they don't understand what the definition of
"it" is. They don't realize what evolution is. They have
come to the problem from artificial intelligence, or systems analysis,
or mathematics, or astronomy, or aerospace engineering. Folks like
Ray
Kurzweil, Bill
Joy and Eric Drexler have raised some alarms, but they are too
dazzled by the complexity and power of human cybersystems, devices
and networks to see it coming. They think the power of our tools
lies in their ever-increasing complexity—but they are wrong.
The biotech folks just don't get it either. People like Craig Venter
and Leroy Hood are too enthralled with the possibilities inherent
in engineering biology to get it. And our "bioethicists,"
like Arthur Kaplan, and those who cling to their human DNA like
it was the Holy Grail or the original tablets of stone, blathering
on like Captain Kirk about what special, sacred things we humans
are—they can't possibly get it. All these people who think
(or fear) that technology will ultimately trump biology have missed
the cosmic point. They are not even wrong. To begin to get it, one
must dispense with artificial boundaries. If you are only thinking
about cybersystems and DNA you can’t possibly get it. And if
you are thinking outside the box, you are still thinking too much
like a human being.
Linus Pauling would have gotten it right away. Erwin Schrödinger
too, and probably Robert Oppenheimer. Bertrand Russell got it. In
fact he named it. What Ray, and Craig, and Eric, and Arthur can't
see is the power of pure chemistry—what Bertrand Russell called
"chemical imperialism." What they don't get is this—
a system does not have to be complex to be transcendently, transformatively
powerful. After all, we and everything we have created are nothing
but the product of "carbon imperialism"—carbon being
the element that all known life is based on. Nothing but the power
of pure chemistry. Living and nonliving materials, everything that
exists in the physical world of our experience burns with that same
electron fire. The fire of the chemical bond.
And Prometheus has returned. His new screen name is nanobiotechnology.
Quick. What's the difference between artificial life and synthetic
biology? Don't know? Neither does anyone else, but that isn't stopping
nanobiotechnology researchers from building them—or it, or
that, or whatever. To stay up to speed, there is always Artificial
Life, the official journal of the International
Society of Artificial Life. According to the editors, the humble
mission of the journal "is [to investigate] the scientific,
engineering, philosophical, and social issues involved in our rapidly
increasing technological ability to synthesize life-like behaviors
from scratch in computers, machines, molecules, and other alternative
media." Whoa!
The federal government is in the game big-time as well. For example,
the Physical Biosciences Division at Lawrence Berkeley National
Laboratory tells us it has established the world's first Synthetic
Biology Department, "to understand and design biological
systems…"
Some people might argue that it is pretty cavalier to work on "artificial
life" or "synthetic biology" before we have even
agreed on definitions for these "things." They might even
point out that "artificial life" containing nonbiological
components or new forms of biology could drastically alter the ecological
balance or even the evolutionary trajectory of life on Earth. Of
course the Lawrence Berkeley folks tell us we "need" synthetic
biology for all kinds of excellent reasons. We need it for the efficient
conversion of waste into energy and sunlight into hydrogen. We need
it to create new life forms to use as "soft" biomaterials
for tissue/organ growth. We need it to spawn new cells that will
swim through the air or water to get to chemical and biological
threats and decontaminate them. We need it, and we will build it,
and it will be OK because we are the good guys (and gals). Our new
life forms will only do good things.
In fact, we are very dangerously confused. To understand how confused,
we must introduce the First Law of Nanobotics: The fusion of
nanotechnology and biotechnology, now called nanobiotechnology,
will result in the complete elimination of the barrier between living
and nonliving materials. In other words, nanobiotechnology not
only has the goal, it has the mandate to break through the "carbon
barrier" of life. The result: We will produce not mere cyborgs,
but true hybrid artificial life forms—or manifestations of
synthetic biology, take your pick. In a previous article on nanomedicine
I described a few of the rudimentary "things" that will
emerge from nanobiotechnology: molecular machines that contain parts
from both the worlds of biology and human engineering. Single-walled
carbon nanotubes linked to DNA. Gold nanoshells linked to antibody
proteins.
But gold nanoshells linked to antibodies are just a simple prototype.
The fact is, we have no idea what artificial life and/or synthetic
biology is, much less what it could do, or how it will behave. A
recent article
in Science provides terrifying evidence of our hubris.
Toward the end of this article, the author explains, "Ethical
and environmental concerns must also be dealt with before synthetic
biology fully matures as a field. MIT, the Venter Institute, and
the Center for Strategic and International Studies in Washington,
D.C., have teamed up to examine issues such as how to keep any new
life forms created under control ... One solution: Alter synthetic
genetic codes such that they are incompatible with natural ones
because there is a mismatch in the gene's coding for amino acids."
In other words, we will be protected because these organisms will
have genomes never before seen on Earth! Perhaps, but that could
also be a description of the ultimate biohazard. If the Ebola virus
is considered a Biosafety Level 4 threat, what level would categorize
a pathogenic organism made completely from synthetic genetic codes?
In order to understand the astonishing leap we are about to make,
one needs to grasp that nanobiotechnology is more than just another
tool. It is also a monumental experiment in molecular evolution
over which we may ultimately have very little control. A nanobiotechnology
device that is smart enough to circulate through the body hunting
viruses or cancer cells is, by definition, smart enough to exchange
information with that human body. This means, under the right conditions,
the "device" could evolve beyond its original function.
Cancer-hunting nanobots are often depicted as tiny robotic machines
—thus reassuringly impervious to fundamental changes brought
on by merging with their biological environment. But they will not
be tiny robots. That mechanical fantasy, promulgated by proponents
of "Drexlerian" nanotechnology who appear devoid of
even the most rudimentary knowledge of chemistry, has been decisively
refuted by people who actually build the components for nanobiotechnology
systems. People like the late Nobel Prize-winning chemist Richard
E. Smalley and the great Harvard bioorganic chemist George Whitesides.
What will really go into our bodies, or out into the environment,
will be hybrid molecular devices composed of both synthetic and
biological components. These "devices" will have been
fabricated to specifically exchange chemical information with biological
or ecological systems. They will not be nanobots, they will be nanobiobots—and
those three letters make all the difference.
In fact, the ability to exchange molecular information with biological
systems will be an absolute requirement for these devices to carry
out the functions for which they will be created. To find cancer
cells, or dissolve arterial plaque, or modify damaged neurological
pathways, nanobiobots will be required to "speak" the
language of biochemistry—our language, evolution's language.
Yet they will not be classifiable as the products of biological
evolution, or genetic or human engineering. They will be true hybrids.
We cannot, must not, assume that our current safety and testing
standards, whether chemical, biological or toxicological, will be
sufficient to predict the behavior of nanobiobots once they are
released into the world.
The precautionary
principle developed for environmental policy states that "where
there are threats of serious or irreversible damage to the environment,
lack of full scientific certainty should not be used as a reason
for postponing cost-effective measures to prevent environmental
degradation." This is generally interpreted to mean that a
lower level of proof of harm can be used in policymaking whenever
the consequences of waiting for higher levels of proof may be very
costly and/or irreversible.
Given that we don't even have definitions for artificial life or
synthetic biology, how would we even begin to apply the precautionary
principle here? But we urgently need to.
Let's take a simple example. Plans are currently underway to create
medical nanobiobots that will use our own metabolic energy (for
example, glucose oxidation) as a source of power. That means these
devices could remain operational as long as we are alive—or
longer if they manage to get into human egg or sperm cells. Any
nanobiobot that develops the ability to propagate in this or any
other manner across even one human generation has fulfilled the
definition of a non-biological life form. A true alien. And it can
happen.
Suppose a glucose-powered nanobiobot has been created to hunt cancer
cells via a component antibody
moiety.
In effect, this nanobiobot has a protein grappling hook designed
to dock it with a specific type of tumor cell. Standard dosing therapy
will require that billions of these nanobiobots be released into
their human "host." If the antibody arm on even one of
these nanobiobots is modified (either by some type of catalytic
recombination with circulating antibodies or by simple chemical
damage) so that it binds to a different type of cell, it could stay
in that body for life, like cryptic viruses such as Epstein-Barr.
If this nanobiobot is modified so that it can attach to a human
sperm or egg cell, it could theoretically stay in the population
for generations.
If this type of nanobiotechnology-based cancer therapy becomes
common (and according to the NCI's nanomedicine site, that is a
real possibility), we could have tens of thousands of people carrying
cryptic nanobiobots. Even though these nanobiobots were designed
for different functions, it is reasonable to assume that they will
have a number of components in common. For example, many of them
may have antibody components that, in turn, have regions of identical
protein structure. These interchangeable parts could act just like
the repetitive
DNA of introns
in eukaryotic
genomes.
What happens when one nanobiobot (say) on a sperm cell meets a second
one on an egg cell? The probability of this is, of course, extremely
low. But if the population of nanobiobots introduced into the body
is high (say, billions), then a one-in-a-million event becomes common.
In fact, microbial and viral systems like E. coli and bacteriophages
enabled the molecular genetics revolution precisely because with
billions (or even trillions) of test organisms in hand, one-in-a-million
events become commonplace.
Suppose in the near future, a routine nanomedical procedure involved
the introduction of billions of nanobiobots designed to scour the
arteries dissolving plaque. Cleaning out the circulatory system
would be considered a "one shot" treatment so that these
therapeutic nanomedical devices (nanobiobots) would not have the
engine necessary to use human metabolic energy as a power source.
But what if, during another "routine" nanomedical procedure,
a second therapeutic nanomedical device (nanobiobot) designed to
vaccinate against cancer is introduced into the same person? This
latter nanobiobot would, by definition, be designed for longevity
so that metabolic energy would likely be the power source. Now,
what if these two meet up and combine, or exhange vital components?
This could happen through physico-chemical damage or perhaps via
some type of catalysis mediated by the host's own complex biochemistry.
Now we have a novel, hybrid nanobiobot capable of crawling through
our circulatory system for life. Or until it exchanges even more
information—either with another nanobiobot or with the body
itself. In the world of biology, this type of event would be called
a mutation.
Even more likely is the "prion"
scenario, in which one of the billions of nanobiobots in the body
is damaged or modified and, as a result, gains the ability to convert
other nanobiobots in a manner that alters longevity, tissue target,
etc. (This is what the abnormally structured proteins called prions
do. Prions are responsible for fatal, mysterious brain-tissue diseases
like "mad cow" and fatal
familial insomnia.) These myriad possibilities bring us to...
The Second Law of Nanobotics: It is not possible to ensure that
devices created using the techniques of nanobiotechnology will only
transmit molecular information to the target system.
This law essentially says it is impossible to ensure that molecular
information only flows in one direction. Just as today's pharmaceuticals
almost always have side effects, there is no natural law that guarantees
against the reverse movement of fundamental chemical information
from the biosystem to the nanobiobot. Any real nanobiotechnology
system—one that uses a combination of biological and synthetic
components—is theoretically vulnerable to a reversal in the
flow of molecular information. This, in turn, will create opportunities
for the unpredictable evolutionary advances of these devices via
a process similar to biological mutation.
Put plainly, if the nanobiobot can modify us there is no way to
ensure that we can't modify the nanobiobot.
Corollary to the Second Law of Nanobotics: Before nanobiobots are
used outside of a controlled research laboratory environment, we
must try to define and understand what it is we are making. And
rigorous algorithms and adversary-analysis systems must be developed
to test these devices to ensure that they are not obviously vulnerable
to the reverse flow of molecular information. Of course, we will
never know this with certainty. But we haven't even started trying
to find out.
What this all means is that within a generation, biology will face
its ultimate identity crisis. Researchers in the field of nanobiotechnology
are racing to achieve the complete molecular integration of living
and nonliving materials. We will hack into the CPU of life in order
to insert new hardware and software. The purpose is to extend the
capabilities of biology far beyond the limits imposed by evolution,
to integrate the incredible biochemistry of life with the equally
spectacular chemistry of nonliving systems like semiconductors and
fiber optics. The idea is to hard-wire biology directly into any
and every part of the nonliving world where it would be to our benefit.
Optoelectronic splices for the vision impaired, micromechanical
valves to restore heart function.
But the moment we close that nano-switch and allow electron current
to flow between living and nonliving matter, we open the nano-door
to new forms of living chemistry—shattering the "carbon
barrier."
This is, without doubt, the most momentous scientific development
since the invention of nuclear weapons. When we open the door and
allow new forms of chemistry to enter, we will change the very definition
of life. Yet no coherent strategy exists to identify the moment
when nanoengineered smart materials cross over into the realm of
living materials. Could we even recognize a noncarbon life form
at the moment of its creation? The answer seems intuitively obvious
until we remember that we too are made of materials. That we too
are machines.
Humans operate entirely on electric current. There are 10 trillion
living
cells in your body, each powered by an electrical potential
of 12,000,000 volts per meter. A thousand times as hot as the plug
on your wall. The voltage of life is produced inside every cell
by a sophisticated electrochemical power generator. Each subcellular
"mitochondrion"
is a protein nanomachine designed by evolution to burn sugar, one
molecule at a time. The heat from this controlled burn yields high-energy
electrons that are the anima of the living state. Every move you
make can be traced back to a specific flicker of this electron fire.
Electromechanical systems drive the contraction of your heart. Electro-optical
systems capture the image on your retina. Layers of electrochemical
switches form the architecture of the neural CPU in your brain.
The bioenergetic transformations that fuel life are an amazing
sequence of reactions that convert light into chemical
bond energy. The biological ecosystem of Earth is one gigantic
solar-powered fuel cell. Plants harvest the sun and animals harvest
the plants. The first step is the light-driven fusion of water and
carbon dioxide into sugar via the photosynthetic organisms—green
plants and some microbes. This sugar is the fuel that drives the
chemical engine of animal life. Our mitochondria use bio-catalytic
converters to strip electrons from sugar and feed them into your
cellular power grid. As electrons move between energy levels, current
flows.
Electronic conduction thus provides the true interface between
living and nonliving materials. Today's technology does not allow
fabrication of components that plug directly into this interface,
but we are getting close. In the early 21st century, nanotechnology
will create the tools to hard-wire into the CPU of life, while biotechnology
will provide a complementary molecular schematic of our living circuits.
It is the engineering destiny of nanobiotechnology to create the
first electro-molecular interface between the living and nonliving
worlds. Or, more correctly, the first interface that does not discriminate
between the living and nonliving states of matter. Fabrication of
the world's first true Biomolecule-to-Material interface will be
infinitely more than a landmark in the evolution of human technology.
Like the separate days of Genesis, the first nanofabricated BTM
interface will be its own monumental act of creation and a crucial
step on the path to bona fide living materials, aka artificial life.
In the history of science, the conduction of signals between living
and nonliving materials will be divided into the pre-nanotech and
nanotech eras. We are still pre-nanotech, which means that a direct
BTM interface has yet to be fabricated, although bioengineering
has created synthetic devices that communicate indirectly with living
materials. Take an artificial
pacemaker. This device transmits an electrical voltage to the
biological pacemaker cells of the heart. In a healthy human, these
pacemaker cells generate their own action potential, an electrical
waveform of about 100 millivolts. This may not sound like much energy
until we remember that this electrical potential is sustained across
an insulating membrane only five nanometers thick. That is 5 billionths
of a meter. So the energy of an action potential is almost 20,000,000
volts per meter. Compare this to the 12,000 volts per meter at a
standard wall plug. Healthy pacemaker cells spark the electrical
wave that drives heart muscle contraction. When these cells malfunction,
an artificial pacemaker may be implanted to take over. Waves of
electrical voltage generated at the metal lead of the artificial
device cross over to living tissue and initiate normal muscle contraction.
While the pacemaker is a magnificent feat of bioengineering, it
does not operate via a true BTM interface. The metal lead of the
artificial pacemaker, a small wire, is physically embedded in cardiac
tissue and the wave of voltage spreads from the charged tip into
the surrounding region. Only pacemaker cells will respond to the
artificial voltage wave by initiating a further action potential.
So the living system must identify the artificial signal and act
upon it. The voltage produced by an implanted pacemaker, like a
radio signal, will pass through space unnoticed unless there is
an antenna to pick it up. In this case the receiving antennae are
individual protein molecules embedded in the membrane of the living
cardiac pacemaker cell. Other heart cells feel the electrical signal,
but do not respond to it. They may be considered as nonspecific
noise in the system. We must flood the local tissue with electricity
in order to obtain the desired response.
This strategy is extremely effective, but it does not constitute
a direct interface between living and nonliving materials. In the
end, the pacemaker does not "know" that the target cells
are out there. It will send its signal regardless of whether it
is received or not. Likewise, the cardiac pacemaker cells do not
"know" that the charged metal lead is out there; they
simply respond to an electrical shock.
By contrast, a nanofabricated pacemaker with a true BTM interface
will feed electrons from an implanted nanoscale device directly
into electron-conducting biomolecules
that are naturally embedded in the membrane of the pacemaker cells.
There will be no noise across this type of interface. Electrons
will only flow if the living and nonliving materials are hard-wired
together. In this sense, the system can be said to have functional
self-awareness: Each side of the BTM interface has an operational
knowledge of the other.
Molecular imprinting offers one nanotechnology strategy to build
a BTM switch in the near future. A molecular imprint works exactly
the way one would think. An isolated biomolecule is surrounded by
some type of self-reactive liquified matrix, often an unpolymerized
plastic like acrylamide. A cross-linking reagent is added, and a
polymer
forms around the biomolecule. When the biomolecule is removed, its
ghostly outline is etched into a surface of solid plastic. The imprint
fits the biological surface with atomic precision so this nanoengineered
component is now a socket into which any identical biomolecule can
be plugged. In the case of a pacemaker, the voltage-sensitive protein
switches from cardiac cells would be imprinted into an electronic
material. The imprinted material would be nanomachined and joined
to an equally small power generator. The entire nanodevice, except
for the imprinted socket, is then coated with a biomimetic ultrathin
film. This coating makes the surface compatible with heart tissue.
This nanopacemaker will occupy less than 1 cubic micrometer, smaller
than a single bacterium. To complete the BTM interface, a living
cardiac pacemaker cell is excised from the patient and plugged into
the socket created by the original molecular imprint process. This
can be accomplished with a micromanipulator similar to those currently
used to move living nuclei in and out of cells. The "hard-wired"
nanopacemaker is implanted into the heart where it is cemented into
place by the body's normal healing process.
The example above was selected because it is relatively simple,
using technology that is already in the pipeline. Far more sophisticated
strategies are on the horizon. One involves literally drawing the
imprinted surface around the biomolecule by polymerizing monomers
with a computer-targeted laser. When bioengineers begin to fabricate
these BTM interfaces we will have entered the nanobiotech era.
If we continue to insist that life on Earth can only result from
biological evolution, then the first BTM interfaces built by nanobiotechnology
will be speciously trivialized as just a great invention of Homo
sapiens. We will congratulate ourselves and conclude that the supremely
gifted toolmaker has built the first portal between the worlds of
living and nonliving materials. This simplistic view of nanobiotechnology
is very much like humanity's current strategy in the search for
extraterrestrial life. In a chemically diverse universe we insist
on a perversely self-congratulatory strategy. Water and organic
molecules, such as methane, are the identified spoor on this trail.
We look for these signs because the biology-centric assumption is
that aliens will be just like us, only very, very different—little
green people with acid for blood, sentient jellyfish with a taste
for cheeseburgers, or insects that have evolved with a sense of
humor. Even search strategies that use "universal mathematical
constants" ignore the possibility, proposed by some postmodern
philosophers
of science, that formal modern mathematics is a function of
cognitive structure unique to humans, or less specifically to a
narrow range of beings similar to humans, for example, hominids.
The point is that technology analysts who can only see life as some
variation on biology will see the BTM interface as a way for "us"
to plug into "it." Within this paradigm there are no consequences
for the definition of life, only new enhancements for the one true
life form: biology. We hold up the mirror of humanity and see our
own image reflected in the universe.
Most dictionaries define biology as "the science of living
things." But the (correctly) limitless nature of that definition
is truncated when plants and animals are immediately used as the
prime examples. NASA, an agency that should know better, has saturated
the media for decades with hypnotic invocations of water and organics
as the true signs of extraterrestrial life. Meanwhile, Hollywood
and pop culture endlessly anthropomorphize
aliens. Robots get the blues. Silicon sentience springs directly
from human mythology. Stories of demonic computers and undead cyber-blood
lust are endlessly refilmed with really cool graphics, a variety
of soundtracks, and excellent eyewear. Skynet, the "self-aware"
computer system of the "Terminator" series, hates us and
wants us dead. The equally demonic cyber-beings of "The Matrix"
want to enslave us and eat our energy (making this computer both
physically dangerous and dangerously ignorant of the physical laws
of the universe). It is distinctly ironic that when we consider
aliens, life on Earth infuses our scientific models, our dreams,
and our entertainment. We could call this "the biology paradox."
The biology paradox makes xenobiology speciously comprehensible,
but by clinging to it we dismiss almost all of the chemistry in
the universe.
It is time for serious students of sentience to accept that common
usage has rendered the term "biology" completely useless
in the nanotech age. Thinking outside the biology box leads to the
alternative, much more radical concept of living materials—materials
with anima.
To describe this new state of life, I suggest a contraction of
the term "anima-materials"—"animats." This
term has previously been used to describe adaptive or cognitive
systems capable of robust action in a dynamic environment. The goal
of these systems involves the creation of higher levels of cognition
from many smaller processes. Many scientists who work in this field
appear ready to dismiss chemical sentience as smaller and simpler
than anything they would consider smart. But we must not assume
that minds are built from mindless stuff. Chemical intelligence
can manifest as the ability to catalyze a single chemical reaction.
It is a dangerous, and possibly terminal, error for the children
of carbon to dismiss the power of pure electron fire. Much of our
fear
of bioterror is based on the power (chemical intelligence) of
a single molecule that allows it to block a single metabolic reaction
inside the human body.
Better to heed Bertrand Russell's prescient warning that "Every
living thing is a sort of imperialist, seeking to transform as much
as possible of its environment into itself." Russell goes on
to use the term "chemical imperialism" as the driving
force for biological life. The obvious corollary to this warning
is that chemical imperialism spawned human intelligence, not the
other way around. Therefore, the definition of an animat as a living
material should have primacy over any definition involving more
complex cognitive functions. If we accept this logic, the creation
of the first BTM interface by nanobiotechnology will require a new
operational definition for the living state.
To expand the chemical franchise of the living state we must first
deconstruct biology. The
Human Genome Project sold us the concept that DNA is the chemical
basis of life. But, in fact, that is not true. DNA is the result
of life, not its cause. Our genetic code is the crowning achievement
of biochemistry, not its progenitor.
It is crucial to keep this distinction in mind when considering
the concept of animats. Life is not defined by DNA but by a continuous
chemical struggle against entropy. The second
law of thermodynamics tells us that all natural systems move
spontaneously toward maximum entropy. By literally assembling itself
from thin air, biological life appears to be the lone exception
to this law. The gaseous molecules snared by plants during photosynthesis
were once free to roam the entire atmosphere of Earth. Plants—Earth's
primary producers —fix gas molecules from the air and minerals
from the water into sugars and proteins. Humans eat the plants,
or we eat the animals that eat the plants. Now those molecules that
were free to roam the skies and waters must be where you are, go
where you go, and do what you do. Clearly, the atoms in your body
have experienced a radical reduction in entropy.
But thermodynamics takes the full measure of the physical world.
What little biology can build is barely visible against the chaotic
horizon generated as the sun exfoliates into space. Like a tiny
windmill in the solar hurricane, the wheel of life is turned by
a unique set of chemical reactions that capture and channel the
least part of that storm of dissipating energy into further cycles
of replication. Biological life is a tiny stowaway on the entropy-powered
craft of our solar system.
Life, then, is not based on DNA but on a chemical programming language
spoken by a discrete set of biomolecules. This language directs
the set of operations necessary to assemble the next generation
of biomolecules. DNA or RNA, the genetic material, stores the directory
of available biochemical operations but does not execute them. The
program steps for replication are executed by a set of protein catalysts
collectively known as enzymes.
It is probable that the first biological life forms were RNA molecules
capable of both catalytic replication and data storage—so-called
ribozymes.
Through evolutionary time, RNA generated two biochemical subroutines,
proteins and DNA, to carry out some of the operations of replication
and data storage with greater efficiency. Yet a cursory look at
the molecular biology of the cell proves that RNA
retains its central role. If life is viewed as a discrete set of
chemical operations, then nanofabricated components that directly
interface biological and materials chemistry must create the possibility
of new life forms. These nanofabricated components are, in fact,
the next generation of self-replicating systems: not enzymes but
animats.
One could argue that it is too early to be talking about animats.
It is easy, and reassuring, to dismiss even the most advanced nanobiotechnology
systems of the near future as mere devices. But if biological evolution
is any guide, that viewpoint is both specious and potentially catastrophic.
During the 3-billion-year operation of the algorithm called evolution,
revolutionary new adaptations often began as trivial events. A small
genetic mutation resulting in a slightly altered protein that provides
an incremental, almost trivial, enhancement to catalytic function.
Thermal tolerance is a classic example. A mutation to the DNA sequence
translates into a modified physical structure for an essential protein.
This new structure has enhanced thermal stability, which means it
retains enzymatic function at a higher temperature than the original.
As a result, the mutant is capable of 100 percent catalytic efficiency
in climates a few degrees hotter than normal. This change in protein
structure will only involve the rearrangement of a few atoms, making
molecular evolution the original nanoengineer.
Over time, the heat-tolerant progeny of the original mutant may
be able to migrate into a warmer climate: say, move down the Sierra
Nevada into Death Valley. But it takes thousands of reproductive
generations or more for this migration to actually occur. The original
mutation will not become essential for a hundred thousand, or even
millions of years. Evolution covers enormous distances one angstrom
at a time, which means it is almost impossible to catch an adaptation
at the exact moment, or even in the exact generation, that it becomes
essential for survival. Likewise, it is highly probable that the
BTM interface will evolve from smart material to living material.
This means that, in order to find the moment when the first animat
appeared on Earth, we will have to backtrack from the future. Or
be watching the present very, very carefully.
Based on this evolutionary model, it is highly unlikely that animats
will spring fully grown upon the Earth. It is much more likely that
animats will initially evolve as part of a larger biological system.
In order to identify the first true manifestation of a living nonbiological
material, we must develop a definitive test to distinguish an organism
that is at least part animat from one that carries a smart material
designed simply to assist or enhance life function.
This brings us to the Third Law of Nanobotics: The carbon barrier
will be eliminated when humans create the first synthetic molecular
device capable of changing the state of a living system via direct,
intentional transfer of specific chemical information from one to
the other.
This law formalizes the concept of animats and leads directly to
the "Animat Test," which is designed to identify the moment
in time when life on Earth evolves to include both biological and
nonbiological materials—the date when we break the carbon barrier.
Let us define a life form as an entity that reduces entropy by
self-executing the minimum set of physical and chemical operations
necessary to sustain the ability to execute functionally equivalent
negentropic
operations indefinitely across time. Given that, a life form will
be considered an animat (living material) if all the information
necessary to execute that minimum set of physical and chemical operations
cannot be stored in DNA or RNA. The corollary: If all the information
necessary to execute that minimum set of physical and chemical operations
can be stored in DNA or RNA, the life form is biological.
In the beginning, nanobiotechnology will create minute supplemental
lifesaving medical devices for humans. The purpose of these devices
will rapidly expand to include the performance-enhancing—an
inexorable development I have discussed
previously. Some of these things will remain devices. But some
will have the potential to evolve and should be termed proto-animats.
The animat test is designed to be a practical engineering tool to
identify the point in time when the proto-animat crosses over and
becomes a true living material, an animat. The conditions of the
test are independent of both the physical structure of the life
form and the physical modality by which the life form perpetuates
a negentropic existence across time. That modality could include
replication, and/or duplication, and/or continuous self-restoration.
The test cannot be applied to entropic life forms since human understanding
of physical laws does not currently allow discrimination between
life forms and other natural phenomena without cycles of entropy
reduction.
Much as we track incoming comets on a possible collision course
with Earth, extraordinary vigilance is required as we transition
into the age of nanobiotechnology. If the evolutionary model prevails,
we are seeking to identify proto-animats: smart materials potentially
capable of evolving into animats, living materials. This, in turn,
will require a radical expansion of our thinking with respect to
the potential sources of artificial life. Up till now (and thanks
to people like Ray, Bill and Eric), most models have focused on
computers and machine intelligence. Smart materials can certainly
contain computers. But it is unlikely that animats will spring to
life via some Hollywood scenario whereby a supercomputer crashes
into A.I. self-awareness and begins photovoltaic-powered reproductive
assembly of little A.I.s (subsequent end-of-the-human-world-as-we-know-it
scenarios optional, heavy metal sound track preferred). If the evolutionary
algorithm is any guide, animats will break the carbon barrier the
way the Bell X-1 broke the sound barrier, carried aloft on the wings
of a mother ship. The mother ship will be named Homo sapiens. The
initial manifestation of an animat life form will be evolutionary
in form, but revolutionary in function. There is also the possibility
of progression from the ternary fusion of biological life, machine
intelligence, and smart materials (proto-animats). But it is crucial
to recognize that living materials need only think with their chemistry.
No Boolean or humanoid logic is required to qualify as life. The
absolute progress of chemical imperialism can only be measured in
entropy reduction.
Unless we know what we are looking for, the first proto-animats
will be invisible in the storm of nanobioengineering systems expected
to come online over the next generation of human life. Most of these
nanodevices will not have the potential to evolve beyond cyborg
mode, i.e., technical augmentations to biological life forms. There
are many future scenarios in which humans will need their machines
to continue to live, but until an animat is carried through time
as part of a life form's self-executing set of essential operations,
the carbon barrier will remain intact. But when the portal between
two worlds is atom-size, how will we know when it finally opens?
In a world where we are already doing research on artificial life,
synthetic biology and nanobiotechnology, this question cannot possibly
be considered academic. Materials will continue to get smarter until
they finally break the carbon barrier. In the near future, some
nanoscale cyborg technology will undoubtedly be designed to propagate
along with the host using molecular self-assembly, the same strategy
used by biological systems.
But self-assembly is not unique to living systems and, therefore,
cannot be used as the litmus test for new forms of life. Water molecules
can self-assemble into the simple crystalline pattern of an ice
cube or the infinite complexity of a snowflake. Quartz and other
inorganic minerals can spontaneously crystallize and grow with a
concomitant reduction in entropy, yet geodes
are definitely not alive.
However, molecular self-assembly is an excellent strategy for building
nanomachines and many researchers are studying ways to harness this
phenomenon. Such nanomachines could even be designed to use self-assembly
to replicate. The original "Grey
Goo" scare (the very mention of which is anathema to most
nanoscientists) involved a scenario whereby endlessly self-replicating
nanomachines literally covered the earth. This scenario is generally
attributed to speculation contained in Eric Drexler’s 1986
book "Engines of Creation."
While the science behind the original Grey Goo scare was and remains
completely unrealistic, we are getting better and better at using
molecular self-assembly to build,
maintain and propagate nanomachines. For example, it is certainly
realistic to posit nanomachines that use ingested trace metals and
semiconductor nanoparticles (for example, silica) to replicate inside
the host's cells, including germ cells. This type of device could
enhance human performance and even move from parent to child, yet
would not be considered to be a new life form (either alone or in
combination with its human host) unless it could pass the animat
test. More to the point, the animat test gives us a way to determine
when a smart material crosses over and becomes a life form.
It is ironic that, because of nanobiotechnology, we have never
been closer to a Grey Goo scenario—although the actual color
will more likely be green or red. Because biomolecules learned self-assembly
through billions of years of evolution, nanobiotechnology has a
tremendous advantage when it comes to applying this particular strategy
to create artificial life.
In fact, we have put into motion research that will create every
component necessary to build an animat. One formula is as simple
as A + B + C.
A = Nanobiotechnology devices that can survive and function inside
human beings. Many therapeutic devices in development for drug delivery,
cancer therapy, etc., are designed to survive in the physicochemical
environment of the body.
B = Nanobiotechnology devices that can derive energy from biological
metabolism. Many nanomedical devices will be powered by the fuel
available inside the human body. A common idea is to take our own
glucose-oxidizing enzymes and use them as a fuel cell for the nanobiobot.
C = Nanobiotechnology devices capable of copying themselves by
molecular self-assembly.
Which creates a completely realistic animat formula. A + B + C
= a self-replicating nanobiobot capable of living inside the human
body powered by our own metabolic energy.
Of course, scientists are not intentionally putting A together
with B and C. No one is trying to create the first true animat—they're
just working on rudimentary forms of artificial life or synthetic
biology. But if, as part of this benign research initiative, they
happen to create nanobiobots some of which have traits A or B or
C—our definition of life will have changed forever.
Does this mean we will immediately cease to be human? Probably
not. The most probable scenario is that an array of proto-animats
will be carried as an evolutionary adaptation that enhances biological
function for generations before any of them become an essential
part of our phenotype. After that...
If the animat test described here is not sufficient, let it stand
as a challenge for the development of a completely rigorous test
for the unequivocal identification of nonbiological life forms.
The larger point is that humanity must initiate a search-and-test
protocol now in order to prepare for the arrival of the literal
alien from within.
Nanofabricated animats may be infinitessimally tiny, but their
electrons will be exactly the same size as ours—and their effect
on human reality will be as immeasurable as the universe. Like an
inverted SETI program, humanity must now look inward, constantly
scanning technology space for animats, or their progenitors. The
first alien life may not come from the stars, but from ourselves.
© 2006 Salon.com. The ideas stated here reflect
the personal views of the author. They are in no way related to
his professional affiliation with Alfred University.
| 
