Drexler Responds

Three weeks ago I critiqued Eric Drexler’s book Radical Abundance. Below the fold is his reply, and my response:

————————-

Hi Robin,

While criticism is great, I was surprised by the specifics of your recent critique of my new book, Radical Abundance. The book centers on prospects for high-throughput atomically precise manufacturing, and because it raises some urgent questions, I would very much like for its message to be understood.

As I see it, your review primarily criticizes me for saying something I didn’t (and wouldn’t) say about material wealth and human satisfaction, and for not explaining what I did in fact explain regarding the likely pace of particular technological advances given a particular set of technological preconditions.

Regarding what I didn’t say, your post states that

Drexler thinks most would feel this new income to be “enough,” and so care little about income differences, and have little reason to conflict…

and in support of this, quotes my statement that

[Particular conditions] would decrease pressures to compete for access to markets and natural resources simply because there can be no vital interest in resources that are no longer scarce or important, nor a vital interest in export markets once imports and trade balances are no longer essential to material well-being. [emphasis added]

I made this remark in the context of international conflict, where “vital [national] interests” are something quite different from individual human desires. Since my vocabulary includes the term “positional goods”, it would be hard for me to make the mistake of thinking that “most would feel this new income to be ‘enough’” (and note that “enough” shouldn’t be in quotes here).

Regarding scenarios that involve a sharp acceleration of progress in the wake of a relatively accessible technology threshold, I’ve already attempted to address the concerns you raise. In particular, to extrapolate from progress to date in “nanotechnology” is to mistake a marketing label for a research program. Work funded under that label has been directed overwhelmingly to materials science, and for peculiar historical reasons that I describe in “A Funny Thing Happened on the Way to the Future… ”, US programs were quietly but deliberately directed away from atomically precise manufacturing. Thus, extrapolations from from past progress in “nanotechnology” are misleading, a bit like trying to project passenger aircraft capabilities from progress in catapult technology.

What is far more relevant is the enormous, largely unrecognized, and hence surprising progress in atomically precise fabrication through research in the molecular sciences. From the time when I wrote Engines of Creation, the scale of three-dimensional atomically precise fabrication has expanded from hundreds of atoms to many millions. Barriers to more rapid exploitation of these capabilities are largely conceptual and institutional.

Your post says that “when [Drexler] moves outside his tech expertise he succumbs to seriously wishful thinking in expecting [what had been called] nanotech to come soon and suddenly”. When I discuss the the pace of progress, I am careful to speak in terms of potential rather than prediction,and I am careful to distinguish between physics-based engineering analysis and speculations regarding future human actions. The potential for sharply accelerating progress, given particular technological preconditions, has a concrete, technical basis that is discussed in the main text and further explored in Appendix II. As for timing of all this (soon?), I don’t recall stating an expectation — again, I only describe potential.

I look forward to seeing you again on your next visit to Oxford, when we can have another go at hashing out some of these important questions.

With my best regards,

— Eric

————————-

My (Robin’s) response:

 

Hey Eric,

I accept that you talked about natural resources no longer being scarce or important in the context of international conflict, though I find it hard to imagine nations not caring about resources if their citizens still care greatly. You also talked about “breaking the link between human development and material economic progress,” which I also found hard to imagine if more developed humans still cared greatly about having more and better stuff.

I also accept that while you did at times suggest nanotech might come soon, you didn’t push this point. But on suddenness I thought you did much more than merely point out the possibility of nanotech appearing suddenly; you consistently relied on that scenario in your analysis of social consequences. I gave many quotes in my post. Yet in your analysis of nanotech, I didn’t see much on why a sudden scenario is likely, beyond it being possible.  You certainly didn’t focus on making clear your arguments on that issue within anything like the energy and clarity you put into arguing that nanotech will eventually be possible.

The main argument you gave for why a nanotech revolution could happen suddenly is that new nanotech designs could “unfold at the speed of new digital media”, i.e., we could sent such designs around fast as digital files. But if this were all that was needed for a technology to improve rapidly we should now see rapid gains in the design of novels, music, and software.  Yet we don’t see such gains in novels or music, and what gains we see in software are likely because of hardware gains, not because we can share software files fast. Similarly, the mere ability to share nanotech designs fast hardly suggests that such designs will improve rapidly.

yours, Robin

GD Star Rating
a WordPress rating system
Tagged as: , , ,
Trackback URL:
  • Daniel Carrier

    “and what gains we see in software are likely because of hardware gains, not because we can share software files fast.”

    In order to have good software, we both need to develop the software, and have the hardware to support it. The fact that the gains we see are correlated with the hardware would suggest that the hardware is the bottleneck, and the gains in software would be even more rapid, if we had the hardware to support it.

  • themgt

    That’s shockingly obtuse. You don’t think the last 20 years have shown progress in music and software, even though I can now store months of music on something that will fit in the palm of my hand and your own links states “a factor of 43,000 was due to improvements in the efficiency of software algorithm”. How long did electricity or running water take to reach the level of global market penetration that smartphones have hit in under a decade?

    • http://newstechnica.com David Gerard

      Yeah, it’s pretty much wrong. As with everything else, you can now know a hell of a lot about music and that informs the music being produced. And that’s without even mentioning the technical advances in the production of music. The mere fact that anyone can record now, whereas this was literally not true thirty years ago.

  • IMASBA

    I wrote this as a comment on Hanson’s original critique:

    “Yes, relative poverty will always motivate people to bring change. People will always demand a fair distribution of everything technology has to offer. But it is not clear from the quote that Drexler is actually saying people won’t care about relative poverty anymore, it sounds more like he’s saying people won’t fight anymore over trade routes or some rare resource that can only be found in some locations because nanotech can work with abundant materials.”

    It seems like I was right: Drexler was talking about international conflict.

    “I find it hard to imagine nations not caring about resources if their citizens still care greatly.”

    Again, it’s about international relations: people in every country can be putting the rich to the guillotine, and still there can be world peace between the countries if no country controls a rare substance or trade route (the nanotech can work with sand, water and air, so to say). This is what Drexler means.

    • VV

      (the nanotech can work with sand, water and air, so to say). This is what Drexler means.

      This looks more like magic than plausible technology.

      • IMASBA

        It’s by manner of speaking. The point Drexler and I were making is that a world with advanced technology, especially nanotech is less likely to depend on a few highly localized natural resources.

      • VV

        That’s far from obvious.

        1) Energy is scarce. Modern technological processes are already quite efficient, there is room for improvement but not so much. Solar, wind, and other forms of distributed energy sources are intrinsically limited. Unless world population is greatly reduced or we settle for much smaller energy consumption per capita (which doesn’t exactly sound like “radical abundance”) we’ll have to rely on things we dig up from the ground: fossil fuels and fissiles.

        2) Even in the most optimistic Drexlerian nanotech scenarios, you can’t transmute elements. You need specific elements to construct certain materials or perform certain chemical reactions. Substitutes might not be competitive. These specific elements might be rare in economically usable forms (think of Coltan).

      • IMASBA

        1) Fossil fuels and fissiles WILL run out, so we have to find something else. A hyper-advanced civilization will have very cheap, highly efficient solar power and nuclear fusion that can easily power the world for many millions of years to come.

        2) Advanced nanotech will allow for a graphene revolution (and probably bio-revolutions of algae and bacteria producing things like plastics for us) and very efficient recycling. Also, the further you go into the future, the more stuff like coltan will be on scrap heaps and in appliances all around the world instead of under the ground in a few countries.

        Naturally some countries will have some advantages over other countries but on the whole the differences will be smaller than they are today.

      • VV

        Fossil fuels and fissiles WILL run out, so we have to find something else.

        This doesn’t mean that the replacement will be as much abundant and cheap.

        A hyper-advanced civilization will have very cheap, highly efficient
        solar power and nuclear fusion that can easily power the world for many
        millions of years to come.

        I’m always wary of cornucopian predictions like that. Are they based on science or mere wishful thinking and blind extrapolation from the historically exceptional industrial revolution?

        By the way, modern photovoltaic cells perform at an efficiency within a factor of two/three of the theoretical maximum (approx 33%). Photosynthesis in biological organisms (that is, natural nanotech) has a maximum theoretical efficiency (sunlight-to-biomass) of approx. 11%, actual efficiency for typical crop plants is 1-2%, while sugarcane and algae can reach 8-9% under ideal conditions.

        Controlled nuclear fusion might never become practical.

        Advanced nanotech will allow for a graphene revolution (and probably
        bio-revolutions of algae and bacteria producing things like plastics for
        us) and very efficient recycling. Also, the further you go into the
        future, the more stuff like coltan will be on scrap heaps and in
        appliances all around the world instead of under the ground in a few
        countries.

        Yes, but that doesn’t seem to imply that the type of cheap, massive atomic level manufacturing Drexler loves to talk about is feasible.

      • Weaver

        Well, even if massive solar (PV and thermal) comes in at 5x the expense of conventionals, it looks just about ok for a generally richer future in the absence of fissile/fossil…?

      • IAMSBA

        Energy will bound the world’s total wealth, you can get somewhere with better organization but in the end you’ll need more energy to “do things”. But here’s the thing, solar power is already cheaper than 5x times the cost of conventionals, it’s more like 1-2x the cost, depending on geography and the inclusion of external costs (pollution, lung disease). There’s also still room for technological advancement and scale increase in solar while fossils become ever more dependent on expensive deep-sea drilling, fracking and grossly underpaid coal miners (imagine the cost of coal if those third world coal miners were paid as much as the people who build solar panels).

      • VV

        (imagine the cost of coal if those third world coal miners were paid as much as the people who build solar panels)

        Uhm, where do you think are solar panels being made and at what working conditions?

      • IAMSBA

        “By the way, modern photovoltaic cells perform at an efficiency within a factor of two/three of the theoretical maximum (approx 33%).”

        You can stack multiple layers that capture light at different wavelength ranges (some of the will also work at night) and get more than 33%, plus if you can just make them cheap enough per m^2 (this is where nanotech comes in), low efficiency doesn’t matter. It’s not unlikely we’ll see highrises and roofs covered in cheap solar foil that has a low efficiency but collectively generates a tremendous amount of power.

        “Yes, but that doesn’t seem to imply that the type of cheap, massive atomic level manufacturing Drexler loves to talk about is feasible.”

        Indeed it does not, but I thought the whole premise here was to see what would happen to a world that does have that kind of technology.

      • VV

        You can stack multiple layers that capture light at different wavelength ranges

        Yes, but they require exotic materials made of rare elements. Anyway, according to Wikipedia, the current efficiency record, using multiple junctions and lens for light concentration, is 43.5%. Given that the Carnot limit is 86%, I doubt actual efficiency is going to improve much more.

        plus if you can just make them cheap enough per m^2 (this is where nanotech comes in), low efficiency doesn’t matter.

        IIUC, solar cell materials are not particularly complex at molecular level, thus I don’t think that nanotech would be of much help.

        Indeed it does not, but I thought the whole premise here was to see what
        would happen to a world that does have that kind of technology.

        Drexler seems to believe that we are headed to a future where each one of us is going to have a nanofactory in their house (or in their smartphone) which can manufacture arbitrary things from cheap goo, little energy and blueprints downloaded from the internet.

        I think that even if atomic manipulation technology advances greatly (which is itself uncertain), the energy and material constraints will still be significant.

      • Weaver

        I agree that in the short-run fossil and fission look like most cost-effective. With enough capital development (and rare earths), there’s a terrestrial solar abundance at theoretical max recovery for something like 10B people at current US consumption levels with a generous 10-factor for waste, hardware limits etc.

        Fissile, properly managed, and accounting for sea water uranium, should give >10K years, depending on how you mix with solar.

        The capital cost of this is all enormous of course, but not unbelievable for a richer future. In the longer run ( > 10K years) there’s fusion or non-terrestrial solar to step for the fossil deficit. But frankly at those timescales I think Robin and Eric would regard all bets as off. ..

  • John_Maxwell_IV

    “what gains we see in software are likely because of hardware gains, not because we can share software files fast”

    What about github, open source, package managers, etc.? By what definition of software progress have they not accelerated software progress?

  • Pingback: Overcoming Bias : Drexler and I Again