Foresight Institute Logo
Image of nano

In-depth analysis of “rift” over nanotech goals

from the required-reading dept.
UIUC mechanical engineering student Jon Horek has produced an excellent study for the IEEE titled A Critical Analysis of National Nanotechnology Research Funding (in pdf). It accurately describes, in some detail, the "rift" between researchers who advocate molecular manufacturing (MM) and those who do not. Horek concludes that the U.S. gov't working group on nanotechnology should increase dialogue with the MM research community. An astute analysis, long overdue.

4 Responses to “In-depth analysis of “rift” over nanotech goals”

  1. Jeffrey Soreff Says:

    assemblers, or positional control?

    I suspect that part of the rift between MNT and non-MNT nanoscience communities could be minimized just by shifting the emphasis in the MNT community slightly, emphasizing positional control more strongly and assemblers less strongly.

    First, "Nanosystems" itself contains a great deal of analysis of mill-type systems, as well as of more programmable robotic systems. In either case we reap the benefits of positional control, e.g. the ability to build complex structures from inexpensive feedstocks.

    Second, there are experimental examples of positional control that we can point to today, most notably enzymes.

    Third, emphasizing positional control allows a much more "gradualist" view of MNT. There are many ways to orient reactants in order to choose a reaction at one chemically equivalent site rather than another:

    • grasping the reactant over much of its surface with an enzyme.
    • blocking part of the reactant's sites with a protecting group
    • grasping the reactant with a programmable manipulator via a covalent handle

    To my mind the interesting question is not "Are assemblers feasible?", but "How well can we control the positions of reactants, and with how much effort?". There is a broad spectrum of positive answers, ranging from

    1. requiring the design and synthesis of novel enzymes, templates, or regiospecific reagents for each reaction
    2. incorporating moeties such as spacer groups (which can be varied systematically, but which require organic synthesis to vary) into catalysts and reagents
    3. incorporating actuators, even if only single DOF actuators, into catalysts and reagents. Note that this does not qualitatively change the situation from one where one covalently binds spacers into tools. The important change lies in shifting the time scale for changing choices from days down to seconds or less.
    4. incorporating systems of actuators with multiple degrees of freedom into tools
    5. or, finally, contructing tools with enough sensors and effectors to perform "closed loop" tasks, such as conditional repetition

    As long as improvements in positional control are pursued far enough that we get positive feedback, improvements in synthetic techniques that give us the ability to improve the positioning tools themselves, then we'll eventually get to full eutactic synthesis.

    If I were evaluating proposals for experimental research directed towards a nanotechnological infrastructure, I'd look at how it improves our ability to control the positions of reactants, or at how it closes the loop and lets the products of positionally controlled synthesis be used to improve the next generation of tools for positional control.

  2. BryanBruns Says:

    Rift or creative tension?

    If we abstract a bit from the personalities and research politics, and put nanotech into the perspective of the history of physics and other science, then the "rift" that Horek describes seems to resemble common tensions between theoreticians and experimentalists. An additional aspect of this case is that molecular nanotechnology theories concern a topic with engineering applications and may be partially tested through simulation modeling. Nevertheless, the underlying, and potentially very creative interaction, is between the tendency of theoreticians to go soaring ahead toward what might be possible while experimentalists want to focus on what can actually be done in a lab.

  3. RobertBradbury Says:

    This is not a problem for a government solution

    I've read Jon Horek's paper and found it interesting because of the apparent assumption that we do not have molecular assemblers. He argues there needs to be more discussion between the government and groups like FI/IMM and possibly funding focused on solving the problems of assemblers and molecular manufacturing.

    I think there is too much emphasis on classical diamondoid assemblers, perhaps because of a desire for better materials properties. The Drexlerian vision can be subdivided into several areas:

    1. Atomic or Molecular Assembly
    2. Assembly methods (Self- vs. Directed-)
    3. Self-Replication
    4. Programmability
    5. Improved Materials (Diamond, Saphire, TiC, etc.)

    A great deal of the nano-Santa vision (that makes serious scientists want to retch) can be achieved with 1-4. We do currently have these capabilities in nature and the biotech industry us an increasing ability to manipulate them in creative ways.

    As pointed out by others, enzymes can do molecular and positional assembly. Existing biotechnology tools can assemble these. A number of profitable biotechnology companies (e.g. Genencor) generate significant income from doing this. The problem is that the nanotechnology vision hasn't caught on in the biotechnology industry yet. This may be because most biologists are still focused primarily on reverse engineering the parts and understanding how they work, instead of assembling useful devices out of them. (I am in the process of writing a business plan for a company I hope will address this problem).

    Steven Block's statements (see abstract), quoted by Horek, are simply incorrect. Biology gives us a variety of examples of how assemblers are built. We are almost at the point using crystallography and protein structure analysis where have a reasonable understanding of how things like DNA polymerase and the Ribosome actually operate. Perhaps Block wants us to focus the funding on continued reverse engineering, rather than trying to construct things such as the Univ. of Washington (Vogel) and Cornell (Montemagno) groups are doing.

    Finally, I will simply note that my questions to representatives from the Navy and NSF at the 1998 and 1999 Foresight MNT conferences lead me to the conclusion that they do not understand enough basic biology to realize how existing self-assembling, self-replicating systems work or what their capabilities are. I think the coordination suggested by Horek will be unnecessary because we should within the next 2-3 years see the construction of synthetic microorganisms that will serve as the foundation for an industry that will provide many of the benefits MNT allows without the requirement for diamondoid nanoassembly. Once the benefits from (1-4 above) become obvious, there will be increasing emphasis on shifting to better materials and increasing the programming flexibility.

  4. MarkGubrud Says:

    Don't widen it

    I don't think it plays to our advantage (those of us who think the assembler concept makes sense and is important) to emphasize this "rift."

    As we saw recently in Science, Eric Drexler's contributions are beginning to receive the recognition they deserve, at least when specific predictions or conceptual innovations can be identified in relation to current developments.
    The big problem for the mainstream scientific community has always been that Drexler's vision, and all of us who have been inspired by it, went way beyond the state of the art or of anything that could reasonably be achieved in the near term. Continually insisting that assemblers and all-purpose nanotech would be quickly achievable if only enough money were thrown at the problem with sufficient direction, only irritates those who already view this kind of thinking as flaky.
    It makes much more sense to talk about large-scale nanosystems and "molecular manufacturing" as long-term prospects that will arise from the current drive in general nanotech as well as other technical trends. The argument ought to be about the ultimate feasibility and achievability of this technology, not about whether we are losing time by not trying to take shortcuts that, in reality, are not immediately accessible.

Leave a Reply