Accelerating Future Transhumanism, AI, nanotech, the Singularity, and extinction risk.


What are Productive Nanosystems?

Nanotechnology is a hypothesized future manufacturing technology which would employ tremendous numbers of tiny robotic arms working together to construct human-scale products. This would come about via planar assembly, where each tiny nanorobot (consisting of perhaps a few million atoms) manufactures a tiny piece of the product, adding it to a main body bit by bit, until something macro-scale is created (for example, a laptop). For a visual example, see the film Productive Nanosystems: from Molecules to Superproducts (warning: file is 86.1MB in size).

How can we make huge numbers of nanorobotic arms? We'd need a reprogrammable nanorobotic arm capable of self-replicating using readily available materials. This will be difficult - at the macro-scale, we've made limited progress with such robotic arms. But if we could create such a self-replicator successfully, then we could instruct it to produce many trillions of copies of itself, and then reprogram those copies to work together to make human-sized products. A full "nanofactory" would require quite a bit of internal complexity. As listed by Eric Drexler, parts required would include:

  • a casing to protect its interior from air, moisture, and dirt
  • inlets for liquid feedstocks to supply molecules for processing
  • molecular sorting mechanisms to purify inputs
  • alignment and binding mechanisms to organize streams of molecules
  • mechanosynthetic devices to process inputs into reactive tools
  • mechanosynthetic devices to apply tools to workpieces
  • mill-style mechanisms to join workpieces into larger blocks
  • programmable mechanisms to join blocks into complex products
  • a port to deliver finished products while protecting the interior space
  • motors to drive moving parts
  • computers to control material flows and assembly mechanisms
  • stored data and programs to direct the computers
  • data communication channels to coordination actions
  • electrical systems to distribute power
  • a cooling system to dissipate waste heat
  • a structural framework to support the casing and internal components
  • How can we possibly build and control robotic arms and other components at this miniscule scale? Using molecular dynamics, we have accurately modeled tiny motors, shafts, gears, bearings, and more. Scientists have already constructed DNA "walkers", "nanocars", molecular circuits, and other basic nanodevices. But putting these components together into something as complex as a nanofactory will take work. Chris Phoenix, Director of Research for CRN, has described some ideas "Design of a Primitive Nanofactory", a breakthrough paper in the field.

    Because of scaling laws, nanofactories will be tremendously productive. Large products, like automobiles, will be fabricated over the course of days or even hours. Large products, like buildings and aircraft carriers, will be built in weeks rather than years. Here is a visual explanation:

    © 2004 Eric Drexler.

    Smaller robotic arms move faster than larger arms. A product constructed by many tiny arms working together will be built much faster than a product constructed by a single, large robotic arm. This is the advantage of greatly improving the functionality per kilogram of tool. Another advantage of nanotech is atomic precision - you can put each atom exactly where you want it. For example, you can cheaply place carbon atoms in a 3-dimensional matrix of covalent bonds. This is called diamond. Because nanomachines would need to be very rigid to operate effectively, most preliminary designs call for diamond. The first nanofactories will likely be capable of building only diamond products, which will be quite effective for many purposes. These nanofactories will probably be desktop-sized, and theoretically capable of manufacturing their own weight in product in a few hours. (The main limitation is heat dissipation.)

    Once you have a reprogrammable nanofactory that builds diamondoid products, you are in business. The blueprints for many common products will be reworked to accommodate all-diamondoid designs, which will be radically more effective than the originals. Stronger, more durable, and so on. This will be of particular interest in military applications, and therein lies the risk. A nanotechnological arms race could lead to our collective demise, or the creation of a global dictatorship. The technological advantage conferred by a successful implementation of nanotechnology would be massive. This technological advantage could be used to greatly expand human freedom or to limit it.

    The Center for Responsible Nanotechnology explains these incredibly complex issues in very clear terms. Safe Utilization of Advanced Nanotechnology is a good place to start. For further exploration, there is the CRN blog, "Thirty Essential Nanotechnology Studies", and Wise-Nano, a collaborative wiki project. Tens or hundreds of thousands of people have had their awareness of the risks and benefits of nanotechnology increased in the past three years due to the actions of CRN. This will increase the probability that the technology is wisely administrated and distributed when it arrives. If you find CRN's work valuable, consider donating.

    The safe implementation of molecular nanotechnology would throw our entire technological and scientific base so far into the future that the quick and effective implementation of SENS would only qualify as a minor subcategory of the potential benefits. Other benefits would include extremely cheap energy, personal transport, manufacturing capabilities, pure water, massively improved agriculture, computing, communications technology... and the list goes on. Conversely, the malicious application of advanced nanotechnology could lead to wars with millions or even billions of deaths.

    Comments (1) Trackbacks (1)
    1. Thanks, Michael! This is a wonderful endorsement of CRN’s efforts, and greatly appreciated. I’ll suggest only one update on the article, which is to replace “convergent assembly” in the nanofactory description with the more recently developed “planar assembly” approach (see Other than that, it’s great!

    Leave a comment