Josh Hall, author of Nanofuture: What’s Next for Nanotechnology, sends this message to Nanodot readers:
Dear Foresight members & friends,
It’s the time of year when many of you are renewing your Foresight memberships, and helping us meet our $30,000 goal for our Challenge Grant by December 31:
http://www.foresight.org/challenge
I believe that the next decade or two will be the period when nanotechnology and AI, along with some of the other technologies of the kind Foresight was founded to watch, really begin to have major effects on the world outside the labs. Here are some thoughts on the subject (this essay is also posted on Nanodot). We hope to expand and deepen the analysis here over the coming year, and we hope you can be a part of it.
Foresight — with Peripheral Vision
Back in the 60s, Marvin Minsky, John McCarthy, and others presided over
a burgeoning field of study, Artificial Intelligence. Using machines that were
pitifully small and underpowered by today’s standards, they made remarkable
strides toward a visionary goal: creating a machine that could think and converse
like a human being.
Then an unfortunate thing happened. In the 70s, the amount of money
going to AI research began to attract political attention, as money will do.
The people not getting the money used political skills to have it redistributed.
The result was that funding shifted from the core of AI to applications — in
fact, the infamous Kefauver Amendment made it illegal for ARPA to fund any
basic research at all! The decade of the 80s was seen as the decade of the
expert system, where techniques developed in AI were used to tackle real-world
problems — but within the field, it was known as “the AI winter.”
What had happened was that in the shift to applications, work shifted to
concerns that were peripheral to the key elements of the original vision. A
machine that plays a good game of chess is not necessarily intelligent. We call
a good human chess player intelligent because the human learned the game by
watching, imitating, modeling, and in general building up a skill. The machine
got the skill by having human programmers figure it out and build it in. It’s the
ability to learn and build skills that constitutes intelligence, not simply having
them. And AI had shifted to be largely a field which built skills directly, instead
of one which studied how to build a machine that could learn them.
Does this sound familiar? It’s essentially the same thing that happened to
nanotechnology two decades later. The core of the generative vision — productive
machinery built to atomic precision — fell out of favor, and was even attacked
by the people who thought they could, or at least claimed they could, produce
the same results by short-circuiting the process.
But of course in both cases the result is something evolutionary instead of
revolutionary. And in both cases, perhaps surprisingly, the missing element is
the same: autogeny.
Walk into any consumer electronics store and you can buy a GPS unit that
would have flabbergasted any AI researcher in the 60s. It knows the map of the
whole continent. It can plan routes, estimate times, and plot your course while
you drive, about as well as a good human navigator — the errors are different
but comparable. It speaks to you in English, and you can speak to it and, for a
limited set of commands, it understands. The GPS seems a tour de force of the
kind of capabilities AIers were trying to build, and it is a damned useful gadget.
But having a GPS won’t help you one single bit when you try to build the next
“AI” device. Neither will a chess-playing program or house-cleaning robot.
Similarly, the products of current-day “nanotechnology” are beginning to
approach, in some respects, some of the possibilities pointed out by Feynman and
Drexler. The density of memory and circuitry is rapidly approaching molecular
scale. An iPod can hold the text of ten tons of books. New materials are being
fabricated whose properties will likely enable single-stage-to-orbit spacecraft.
But while both are damned useful gadgets, neither is going to help you build a
cell repair machine.
But evolutionary advance, along with general scientific and technological
progress, ultimately lays the groundwork for enough new capabilities for
autogenous systems to be built. One can’t be certain, and wishcasting is always a
pitfall to guard against, but there seems to be some movement back to the
center. In AI, Marvin Minsky could say “AI has been brain-dead since the 70s” and
then be invited to keynote the leading AI conference. Ben Goertzel, originator
of the term “artificial general intelligence” and leading proponent of a return to
AI’s roots, reports that he is no longer laughed off the stage at mainstream AI
meetings. There is a AAAI-sanctioned AGI conference series now going in to
its second year.
In nanotechnology, the cracks in the glass ceiling are appearing in the form
of the Battelle/Foresight Roadmap for Productive Nanosystems and some grant
funding for mechanosynthesis work.
I’ll go out on a limb and say I expect the loop to close in AI sometime in the
next decade and in nanotech in the decade after that. The world will become
an interesting place. To foresee with even a cloudy lens, we need to look at
a variety of technologies that could support an autogenous feedback loop and
thus have a revolutionary impact. Here are some candidates:
• Biotech is already built on an autogenous base, the reproductive capacity
of life.
• Software likewise is a substrate capable of supporting autogeny at a
number of levels short of true AI. The intersection of software, datacomm, and
human-based memetics over the past couple of decades has been explosive.
• Robotics: replacing human workers in physical factories, particularly ones
that made humanoid worker robots, would take humans out of the productive
loop, enabling a takeoff.
• Desktop fabricators: the same idea in a smaller package. I expect the
2010s to be a decade of experimentation with them the way the 80s were
for PCs. There seems to be a fairly straightforward path from fabs to
nanofactories, with increased value added at each stage.
The convergence of these, along with AI and nanotech and who knows how
many others I haven’t thought of, will form the core of technological capability
in the twenty-first century. I suspect that a study of the properties of general
autogenous systems will be invaluable in understanding it.
Once this takes hold, virtually everything will begin changing at Moore’s
Law rates. Let’s hope we have enough foresight that the changes will be
improvements.
————————————————————————
Comments are welcome — you can email me at josh@foresight.org, or go to our blog Nanodot (http://foresight.org/nanodot) and respond in the comments field for “Foresight — with Peripheral Vision.”
Please, if you can, chip in on the Challenge Grant at http://www.foresight.org/challenge. Every dollar you donate will be automatically doubled, so Foresight can do twice as much to influence our future in the positive direction that we all hope for.
Josh Hall
josh@foresight.org
Foresight Institute
1455 Adams Drive, Suite 2160
Menlo Park, CA 94025 USA
Tel +1.650.289.0860
Fax +1.650.289.0863
A nanotech technique that can coat any number of common fabrics with a layer of silicone nanofilaments appears ready to produce durable, completely waterproof clothing. From News Scientist, written by Jon Evans “Nanotech clothing fabric ‘never gets wet’“:
If you were to soak even your best raincoat underwater for two months it would be wet through at the end of the experience. But a new waterproof material developed by Swiss chemists would be as dry as the day it went in.
Lead researcher Stefan Seeger at the University of Zurich says the fabric, made from polyester fibres coated with millions of tiny silicone filaments, is the most water-repellent clothing-appropriate material ever created.
Drops of water stay as spherical balls on top of the fabric … and a sheet of the material need only be tilted by 2 degrees from horizontal for them to roll off like marbles. A jet of water bounces off the fabric without leaving a trace ….
The secret to this incredible water resistance is the layer of silicone nanofilaments, which are highly chemically hydrophobic. The spiky structure of the 40-nanometre-wide filaments strengthens that effect, to create a coating that prevents water droplets from soaking through the coating to the polyester fibres underneath.
“The combination of the hydrophobic surface chemistry and the nanostructure of the coating results in the super-hydrophobic effect,” Seeger explained to New Scientist. “The water comes to rest on the top of the nanofilaments like a fakir sitting on a bed of nails,” he says.
The research was published in Advanced Functional Materials [abstract]
—Jim
Michael Berger at Nanowerk has compared the centralized strategy of Russia’s new nanotechnology program with the national nanotech strategies of other countries. From “Russia’s nanotechnology crash program“:
This week’s successful international nanotechnology forum Rusnanotech in Moscow has put a spotlight on Russia’s ambitions to catch up with the leading nanotechnology nations. While Russia has the money, the political will, and a well educated scientific base to be a leading player, it has completely missed the boat on developing its nanoscience programs and nanotechnology infrastructure.
In terms of gross domestic product (GDP), Russia ranks as the eleventh largest economy in the world. But while many smaller countries such as Australia or South Korea, not to mention all of the bigger nations, have invested steadily and broadly in all areas of nanosciences and nanotechnologies for years now, Russia has had no coordinated science policy, no industrial policy, and no commercial industrial base to develop its nanotechnology capabilities. Until last year, that is. In April 2007, the Russian president signed off on a public policy paper that ordered a multi-billion dollar program to develop a world-class Russian nanotechnology industry by 2015.
After describing the Russian program in some detail, Berger concludes:
It is obvious that Russia has chosen a much more centralized approach to developing a nanotechnology industry than most other industrial nations. While this could be a result of the countries past history of large state monopolies, it could on the other hand be the only realistic way of pulling off the crash development of the Russian nanoindustry.
…Time will show if the Russian approach works. With the flurry of deals, projects and cooperations announced over the past few months, not to mention the glitzy Rusnanotech event, they appear to be off to a good start.
—Jim
Nanowires that are superconducting above liquid nitrogen temperature have been produced for the first time, and their properties vary according to the diameter of the nanowires. The fact that they are superconducting over wider temperature ranges than other nanowires may make them uniquely suitable for some nanotech applications. From Laura Mgrdichian at PhysOrg.com “High-Temp Superconducting Nanowire System is First of its Kind“:
Scientists from the California Institute of Technology have, for the first time, created an array of nanowires that are superconducting at relatively high temperatures. This work, published recently in Nano Letters [abstract], could lead to the incorporation of superconducting nanowires into emerging nanotechnologies.
Researchers around the world have been working to create superconducting nanowires, but few studies have investigated the feasibility of nanowires made of high-critical-temperature (high-Tc) superconducting materials and, prior to this work, no such nanowires had been produced.
“We did not know whether high-Tc superconductivity could be maintained for very long, thin nanowires or whether their behavior would follow traditional superconductor models,” said CalTech chemist James Heath [one of the winners of the 2000 Foresight Nanotech Institute Feynman Prize in the Experimental category], the paper’s corresponding author, to PhysOrg.com.
He added, “High-Tc materials as nanostructures are also notoriously difficult to fabricate because their crystal structure and other properties must be maintained, and this is not easy to do.”
High-Tc materials are often more desirable than traditional superconductors because they function at temperatures “warmer” than the boiling point of liquid nitrogen, or 77 degrees Kelvin (about -321 degrees Fahrenheit). While still ultra-cold by everyday standards, these temperatures are easier to achieve in the laboratory because liquid nitrogen, which is commercially available and relatively easy to work with, can be used to cool them. This makes high-Tc materials more suitable for many applications.
—Jim
Using a nanotech drug delivery method to target cancer cells is much more effective than using the drug by itself. In laboratory tests, nanoparticles that include a small molecule of nucleic acid that binds to a target molecule on prostate cancer cells were used to carry a lethal dose of the drug into the cancer cells without affecting cells lacking the cancer-specific target. From the National Cancer Institute’s Alliance for Nanotechnology in Cancer “Targeted Nanoparticles Boost Platinum-Based Anticancer Therapy“:
A research team from the Massachusetts Institute of Technology (MIT)-Harvard Center for Nanotechnology Excellence has custom-designed nanoparticles that can deliver the anticancer drug cisplatin specifically to prostate cancer cells. The nanoparticles are composed of two different polymers and are decorated with a nucleic acid aptamer that binds to the tumor marker prostate-specific membrane antigen. This aptamer ensures that the nanoparticles deliver their payload only to prostate cancer cells.
Stephen Lippard, Ph.D., and Robert Langer, Ph.D., MIT, and Omid Farokhzad, M.D., Harvard Medical School, led the team that developed this new formation of cisplatin. The investigators published their results in the Proceedings of the National Academy of Sciences of the United States of America [abstract].
To construct a stable nanoparticle that would only release its toxic cargo inside tumor cells, the investigators synthesized a modified version of cisplatin that includes a long hydrocarbon chain. As the nanoparticle forms, the hydrocarbon chain associates strongly with the hydrophobic chains of the polymer that forms the nanoparticle’s core. The researchers note that the hydrocarbon chain they chose optimizes both drug encapsulation and drug release inside tumor cells. Once the nanoparticle enters the cell, the modified drug is converted into its active form as a result of chemical conditions inside the cell.
Tests with human cancer cells growing in culture showed that these nanoparticles were taken up specifically by tumor cells and not by healthy cells. Nanoparticles lacking the targeting aptamer were not taken up either. These tests also demonstrated that the nanoparticles release their cargo over the course of 60 hours, providing a sustained lethal level of the drug inside the targeted cells. In addition, the nanoparticle formulation was approximately 100 times more effective at killing tumor cells than was cisplatin by itself.
—Jim
It is with great sadness that we report the death of Prof. Arthur Kantrowitz, a founding Advisor of Foresight Institute and an early supporter of molecular nanotechnology concepts when they were first developed at MIT in the late 1970s by then-student K. Eric Drexler.
Arthur was an amazingly innovative scientist and technologist, as described in his Wikipedia entry, which is worth reading. The fact forums described in Engines of Creation were an extension of Arthur’s Science Court concept.
To be inspired, read his essay The Weapon of Openness which guides Foresight’s work today, especially in the area of Open Source Sensing. We’ll miss his wisdom.
It’s hard to say goodbye to this wonderful long-time friend. —Chris Peterson
A nanostructure called a “gyroid” self-assembled from diblock copolymers provides a scaffold with the proper dimensions so that when light knocks electrons loose from a dye molecule, the electrons and the holes left behind can be separated to obtain an electric current, providing the basis for a more efficient, inexpensive nanotech solar cell. PhysOrg.com points us to this Cornell University news article by Bill Steele, “Nanomanufactured polymer film could lead to lower-cost solar cells“:
You never know where basic research may lead. For decades materials scientists have been experimenting with a corkscrew-like polymer structure called a gyroid. Now an international team of researchers has shown that the gyroid structure can be used to “self-assemble” a low-cost photovoltaic cell.
The idea could lead to more economical solar collectors and more efficient fuel cells.
The prototype is a variation on the Graetzel solar cell, which uses an organic dye sandwiched between two conductors. Forming the conductors into an interlocking corkscrew allows current to be transported out efficiently. The team’s first cell, made in a thin film 400 nanometers thick, has a conversion efficiency ranging from 0.7 to 1.7 percent — low compared with silicon-based photocells, but “amazing” for such a thin film, said Ulrich Wiesner, the Spencer T. Olin Professor of Materials Science and Engineering at Cornell.
“The next step is to make it thicker” so more of the light falling on it can be captured, he said. “We hope that we will eventually get efficiencies that rival silicon-based devices.” Currently available silicon-based solar cells convert about 15 percent of the energy of the light falling on them to electricity, although some new designs promise higher efficiencies.
The work by Wiesner and scientists at the Universities of Cambridge and Oxford in Britain, the Freiburg Institute for Advanced Studies in Germany, Institute Curie in France and the University of Minnesota, Minneapolis, is described in the online version of the journal Nano Letters [abstract] ….
…the researchers assembled a copolymer gyroid film, then dissolved away just the corkscrew part of the structure, leaving a corkscrew-shaped mold that they filled with titanium oxide. Heating then burned off the other polymer part and crystallized the titanium oxide into a solid structure that conducts electrons. This was coated with a light-sensitive dye, and finally the space around it was filled with a material that conducts “holes” (positive charges).
When light strikes the dye it knocks loose electrons, which flow into the titanium oxide framework, while the holes left behind flow into the other conductor. Electrodes above and below the film carry off the resulting current.
The secret of a solar cell, Wiesner explained, is that the electron-hole pairs must find the interface between the two conductors within 10 nanometers (about the width of 30 atoms) so they can separate and flow away, or they will recombine.
“This is why block copolymers are exciting,” Wiesner said, “because that is the characteristic length scale of separation of the two blocks.”
—Jim
To the list of the amazing properties of carbon nanotubes has been added the ability to make nanotech loudspeakers that produce sound without mechanical movement. From nanotechweb.org, written by James Tyrrell (requires free registration): “CNT loudspeaker rips up the design book“:
A transparent carbon nanotube (CNT) thin film developed by Lin Xiao and colleagues at Tsinghua University, China, could turn out to be a wonder material for makers of audio visual devices. The see-through structure emits sound when hooked up to an electrical signal and can be stretched over a display to play audio content … eliminating the need for conventional loudspeakers or headphones.
To emit sound, the device relies on the so-called thermoacoustic effect. In other words, it is the thermal expansion and contraction of air in the vicinity of the thin film (due to the periodic heating of the CNTs) that produces sound, not the mechanical movement of the thin film itself.
…”We can batch-synthesize super-aligned CNT arrays onto 4 inch silicon wafers, which each provide enough material for a continuous thin film that can be up to 60 m long and up to 10 cm wide,” the researchers told nanotechweb.org. “Because the thin film itself does not vibrate, the loudspeaker will continue to work even if part of the film is broken or if the device is mounted on soft materials such as flags or clothing.”
The research was published in Nano Letters (abstract). The researchers also provide videos of the CNT loudspeaker working. The two links included in the nanotechweb.org article did not work for me, but the Nano Letters abstract includes the following links under “Available Supporting Information for This Article” that did work, and I found the sound to be quite respectable.
QuickTime Video
Microsoft Video (AVI)
—Jim
One group of researchers has developed a method to chemically coat single-walled carbon nanotubes (SWNTs) to produce a nanotech assay for trace levels of proteins associated with cancer that is a thousand fold more sensitive than are current assays. A second group developed a mathematical method to permit analysis of samples for several proteins at the same time. From the National Cancer Institute’s Alliance for Nanotechnology in Cancer “Carbon Nanotubes Improve Protein Array Detection Limits“:
To detect cancer as early as possible, dozens of research groups are developing methods to detect trace levels of cancer-related proteins and genes in blood or other biological samples. Those efforts should get a boost thanks to new research results showing that carbon nanotubes can serve as incredibly sensitive optical labels for use in a wide variety of assay systems.
Reporting its work in the journal Nature Biotechnology [abstract], a research team headed by Hongjie Dai, Ph.D., Stanford University and the Center for Cancer Nanotechnology Excellence Focused on Therapeutic Response, describes a new type of coating developed specifically for attaching any number of different types of targeting agents to the surface of single-walled carbon nanotubes. This coating, a branched form of the biocompatible polymer poly(ethylene glycol) (PEG), enabled the investigators to readily couple antibodies to carbon nanotubes. In the experiments reported in their current paper, the antibodies were designed to identify specific proteins immobilized on a standard protein array microchip.
Carbon nanotubes can function as bright Raman optical tags that are readily detected when irradiated with light. Experiments comparing the lower limits of protein detection using an antibody-labeled carbon nanotube tag and a standard fluorescence tag showed that the carbon nanotube-enabled assay was at least 1,000 times more sensitive than the fluorescence assay. At least part of this improvement resulted from the almost total elimination of background fluorescence that can confound other detection schemes. In addition, the investigators found that the Raman tags were useful over a larger range of concentrations, ranging from 10 nanomoles to 1 femtomoles. The investigators note in their paper that the coating they developed also should enable them to create Raman tags that can detect nucleic acids and other types of biomolecules.
Meanwhile, a second group of investigators, led by Beatrice Knudsen, M.D., Ph.D., Fred Hutchinson Cancer Research Center, and Selena Chan, Ph.D., Intel Corporation, has developed a mathematical technique for analyzing the specific spectral output of different Raman probes, making it possible to create highly multiplexed assays using these probes. Unlike traditional fluorescent labels that typically absorb and emit light in a very narrow band of frequencies, Raman probes generate complex frequency spectra that are chock-full of information.
The Knudsen-Chan team, which published its results in the journal ACS Nano [abstract], developed a method for sorting out the various spectral peaks associated with individual nanoscale Raman probes that were part of a mixture of these probes. Each probe was designed to bind to a different biomolecule. In one experiment, the investigators were able to decipher a complex Raman spectrum that included the optical emission from three different Raman probes and thereby determine the amount of each probe in the mixture. The researchers note that their method for spectral analysis is exceedingly simple to conduct and is amenable to high-throughput analysis in any type of multiplexed assay system.
—Jim
New nanotech applications may be made possible by the demonstration of a force generated from light that differs from the more familiar radiation pressure, and that is more versatile because it does not require a reflective surface. This force can be used to make light drive nanoscale machinery on a silicon chip. From Yale University, via AAAS EurekAlert “Harnessing light to drive nanomachines“
…a team led by researchers at the Yale School of Engineering & Applied Science has shown that the force of light indeed can be harnessed to drive machines — when the process is scaled to nano-proportions.
Their work opens the door to a new class of semiconductor devices that are operated by the force of light. They envision a future where this process powers quantum information processing and sensing devices, as well as telecommunications that run at ultra-high speed and consume little power.
The research, appearing in the November 27 issue of Nature [abstract], demonstrates a marriage of two emerging fields of research — nanophotonics and nanomechanics — which makes possible the extreme miniaturization of optics and mechanics on a silicon chip.…
“While the force of light is far too weak for us to feel in everyday life, we have found that it can be harnessed and used at the nanoscale,” said team leader Hong Tang, assistant professor at Yale. “Our work demonstrates the advantage of using nano-objects as “targets” for the force of light — using devices that … match the size of today’s typical transistors.”
Until now light has only been used to maneuver single tiny objects with a focused laser beam — a technique called “optical tweezers.” Postdoctoral scientist and lead author, Mo Li noted, “Instead of moving particles with light, now we integrate everything on a chip and move a semiconductor device.”
“When researchers talk about optical forces, they are generally referring to the radiation pressure light applies in the direction of the flow of light,” said Tang. “The new force we have investigated actually kicks out to the side of that light flow.”
While this new optical force was predicted by several theories, the proof required state-of-the-art nanophotonics to confine light with ultra-high intensity within nanoscale photonic wires. The researchers showed that when the concentrated light was guided through a nanoscale mechanical device, significant light force could be generated — enough, in fact, to operate nanoscale machinery on a silicon chip.
—Jim
