Friday, April 29, 2005

Engines of Distortion


Dear Howard,

It is a pity that Lord Broers has joined the parade of denialists who discuss distorted versions of obsolete scenarios to the exclusion of all research in the field since 1990. The Royal Society report twice misstates my views, citing publications in which I state the opposite.

For those interested in a brief summary, I recommend my essay for the AAAS, "The Future of Nanotechnology: Molecular Manufacturing". For more technical information, I recommend the discussion and bibliographies my website, e-drexler.com.

K. Eric Drexler

Backgrounder
We've seen the nano enemy and they is us
UK misses chance to defuse nanotox issue
Molecular manufacturing: Who needs it, and why?

14 comments:

Philip Moriarty said...

Eric,

[Universal] assemblers "...will be able to use as "tools" almost any of the reactive molecules used by chemists - but they will wield them with the precision of programmed machines. They will be able to bond atoms together in virtually any stable pattern, adding a few at a time to the surface of a workpiece until a complex structure is complete." [Drexler, "Engines of Creation"]

Is this a distortion, straw man, or "obsolete scenario" or do you stick by your vision that molecular manufacturing will be capable of fabricating almost any structure from almost any material? [Recent debate on this topic is here ].

As regards, Broers' statement: "we do not have any idea how to make an autonomous self-replicating machine at any scale" . I don't know about "at any scale" but certainly at the nanometre scale, using the molecular manufacturing concept laid out in Nanosystems , there has not yet been an experimentally viable proposal to implement mechanosynthesis. To me, this qualifies as not having "any idea" of how to build an autonomous self-replicating machine.

Indeed, there is not a single example in the literature to date of the implementation of even one mechanosynthesis step using a molecular tool as described in Nanosystems . Importantly, this absence is not a question of unavailability of appropriate technology (as you repeatedly suggest via references to the moon landing programme). You state clearly in Chapter 16 of "Nanosystems" that Merkle et al.'s ultrahigh vacuum-based diamond mechanosynthesis strategy represents an alternative and viable route towards Stage 1 of molecular manufacturing. A simple low bandwidth (i.e. 1 reaction a day/week) proof-of-principle experiment to demonstrate the viability of the simplest mechanosynthesis tool (the hydrogen 'abstracter') is 'attemptable' using today's technology. Why do you think that a research group has not taken the mechanosynthesis ideas outlined in Nanosystems and tried to implement them in the lab? Could it be that they have "no idea" of how to implement the machine language of molecular manufacturing? Merkle and Freitas certainly appreciate the key importance of putting forward a credible proposal for appropriate mechanosynthesis tools (see here ). Why would this be necessary if the answers were in Nanosystems ? I reiterate: we have no idea how to implement even the simplest mechanosynthesis steps. (Notwithstanding the keen problems with, for example, steric hindrance that preclude the implementation of many of the suggested mechanosynthesis reactions put forward in "Nanosystems" .)

There is a long line of 'denialists' because the concept of being able to build "virtually anything" via mechanosynthesis is clearly misguided. Why did you focus so heavily on H-passivated diamond surfaces in "Nanosystems"? Could it be that, as stated before , “far from delivering the ability to synthesise ‘most arrangements of atoms that are consistent with physical law’ or to manufacture “almost any… product ….that is consistent with physical and chemical law”, an extremely judicious choice of materials system, possible intermediate/ transition states, diffusion barriers, and symmetry is required to attempt even the initial, most basic and faltering steps in molecular manufacturing. “" (I also note that Don Eigler - the world's leading expert in single atom/molecule manipulation - has recently joined that long line of 'denialists'.)

All the answers are not in Nanosystems - in fact very few of the answers are in Nanosystems . The mechanosynthesis concepts laid out there are genuinely fascinating and extremely thought-provoking but: (i) they can only ever - at best - apply to an extremely limited materials set, and (ii) they are associated with immense (perhaps insurmountable) practical difficulties. (You include well-reasoned arguments in Section 6.4.6 of Nanosystems on the stability of surfaces. Why do you seemingly abandon those arguments to argue that 'virtually anything' may be synthesised via molecular manufacturing?) It will require the expertise of the world-leaders in atomic and molecular manipulation to attempt basic mechanosynthesis steps (and, indeed, perhaps there is some scope for interaction with Freitas and Merkle's proposed theoretical research programme in this regard). Simply labelling those who criticise (or critique) your work as 'denialists' serves only to drive a deeper wedge between the MNT and 'conventional' nanotechnology communities.

Best wishes,

Philip

Philip Moriarty said...

A clarification:

"To me, this qualifies as not having "any idea" of how to build an autonomous self-replicating machine"

should be:

"To me, this qualifies as not having "any idea" of how to build an autonomous self-replicating machine using the MNT strategy outlined in Nanosystems".

(There are, of course, other strategies related to the construction of autonomous self-replicating machines (e.g. artificial cells) which have nothing to do with the concepts put forward in Nanosystems .)

Philip

attobuoy said...

Wake up, guys, and smell the coffee. See http://parts.mit.edu

Attobuoy

Anonymous said...

And happy belated birthday, Eric! It is my opinion that you should have a happy one, although I have no idea how you would do it. Maybe cake. Surely someone can make cake. It's not impossible like tabletop fusion or anything. Oh, wait... Anyway, it hasn't been demonstrated conclusively yet that you cannot have a happy birthday, although it seems many would deny the possibility that you could have one because I've not seen you have your happy birthday, which is, actually, enough to convince me that you can't, even though I'm not going to spend the money to fly out and see for myself because someone else is bound to do that. I wonder if the peer review process will be allowed to decide if you can have one or not, or if more feral metadata and educated opinion is enough to permanently soak the candles and deter others from developing an envisioned happy birthday.

Yes or no, nail-it or fail-it, here's to the peer review process which, despite all the strong opinions and passionate arguments you all read in nanobot, is the place where the research pushes on opinion and opinion falls down. I think Eric's got it. Others think not. I'd like to know how God decides which basketball team wins when both teams pray for a victory.

Cheers,
Damian Allis

p.s. hi Jane... er, attobuoy! Until we get the Vioxx thing figured out or someone demonstrates that we convincingly DO know what the hell's going on down there, I'll stick with http://www.sigmaaldrich.com/ as a parts library.

Philip Moriarty said...

Damian,

Enjoyed your comment - sarcasm has always been, in my opinion, one of the higher forms of wit! But, existentialist ponderings on the nature of Drexler's birthday party aside, as you suggest, it all boils down to science & research and not opinion.

Peer review...is the place where the research pushes on opinion and opinion falls down. But the proponents of MNT have yet to put forward a coherent (experimental) proposal for peer review! Furthermore, the vast amount of research that many groups around the world have done to date involving issues of key significance to MNT (surface science, single atom/molecule manipulation, the entire field of chemistry...) has shown that Drexler's vision of being able to synthesise virtually anything is wrong. This, however, does not rule out the possibility of mechanosynthesis being implemented in a a narrow range of covalently bound solids.

Philip

PS Attobuoy's comment re. biological 'nanomachines' is spot on.

Anonymous said...

Oh, I didn't think I was being sarcastic.

Your argument is that there is no coherent (experimental) proposal YET. I appreciate the parenthetical qualification. As a trained theoretician, I take that to mean that we’re off the ground floor. A lack of an experimental proposal is not a convincing argument against the possibility and NEVER has been in any field of science where work is being done. Some of us are actively involved in development and filling in blanks along the way, and that should be good enough for you, a trained scientist, to stop putting forth the same criticism given, I hope, your own embryonic long-term goals as a researcher for your work. It is a shame that even our grant providers aren’t looking past their noses. That’s the beauty of theory. As long as I can afford to pay my electric bill, I’ll be publishing.

You are no more relevant in your presented argument against MNT than a professor who rides the back of a graduate student to finish writing a paper when data is still analyzed. No, the work is not done and the last publication has not been handed in. I’m sorry you weren’t born into an era where “yes” or “no” had been voted on already. Please be patient. This isn’t rocket science.

Damian Allis

Howard Lovy said...

Sherlock Holmes had his archenemy Moriarity, and now Inspector Damian Allis has his own.

Just stepping in for a second to make an observation for NanoBot critics who think I went off the deep end when I branched out from straight business-news reporting and editing: There are very few, if any, places the molecular manufacturing "community" can go to make their voices heard (and heard in their own voices, rather than their critics' interpretations of what they believe) without being subject to ridicule and other forms of arrogant snorting.

So, I'm happy to provide for them a forum that also has the attention of the the nanocommerce crowd. The environmental activist community is treated with the same respect here, even though I strongly believe they're misguided on the nanotech issue.

A publication where everybody has a chance to be heard, where nobody is left out in the cold on the whim of an editor? Can't think of anything more objective than that.

Just one thing I want to ask of anybody who contributes to my "comments" sections. Unlike most blogs, NanoBot does not ghettoize reader contributions into a separate section. I keep it all on the "front page." So all of you are my unpaid stringers and correspondents. Congratulations. Remember that I try to make nanotech understandable to a general audience. So, no lapses into into arcane language or concepts that can be decoded only by other scientists. Challenge yourself to put it in language that invites readers from high school on up to follow your point.

Oh, and unlike this windy note, try not to go too long. Thanks, folks,

Howard

Anonymous said...

Sorry, Howard.

Hi, Philip. Thank you for putting "experimental" in parentheses.

I think "archenemy" is a little too harsh until we've begun chasing each other in derigibles, although I propose that neither of us ever vacation in Meiringen (sorry Philip, had to do it. On the bright side, people probably don't mistake you for the devil incarnate).

back to the grind,
Damian

Philip Moriarty said...

[Howard, very good points re. lapses into arcane language and verbosity. I'll try to keep it more focussed from now on (although these are important issues that warrant extended debate)...]

Damian,

"A lack of an experimental proposal is not a convincing argument against the possibility and NEVER has been in any field of science where work is being done."

Damian, this entirely misses the point and is very much not my argument. Yes, I have stated consistently throughout my discussions/debates with proponents of MNT that experimental work is absolutely essential - how could anyone argue otherwise? However, we need to separate issues of technical/experimental feasibility from impossibility.

1. Technical/experimental feasibility : Claims that a full-blown MNT technology will almost certainly be with us within 15 years (see, for example, here ) abound (Drexler is a little more circumspect in his estimates but stated in 2001 that the technology will be with us in "1 to 3 decades"). To date, only one proposal to implement basic mechanosynthesis steps in the lab (Freitas et al.'s proposal ) has been put forward. This has been debated/peer reviewed at some length and found to be lacking in key areas.

In addition, world-leading experimentalists with a wealth of experience in the practical difficulties of moving individual atoms and molecules around (e.g. Eigler) have keen difficulties with the MNT strategy. Do you not think that it is both closed-minded and somewhat arrogant to dismiss the expertise of experimentalists such as Eigler? As an experimentalist, I am more than willing to listen to the theoretical MNT community and have read Merkle, Freitas et al.'s papers in some depth (see here ). Indeed, as I've mentioned to Chris Phoenix previously, I spent a considerable amount of time thinking about how one might implement Freitas et al.'s scheme in the lab.

I genuinely believe that, for a very small subset of extremely well-chosen materials systems (see below, and previous posts above), computer-controlled chemistry at the atomic/molecular scale is a fascinating and exciting area of research. I am therefore more than willing to collaborate with Freitas and others on problems of direct relevance to MNT (and have said so in the past). (Note also that it's not a matter of riding the back of a graduate student to finish writing a paper when data is [sic] still analyzed (something I hope that I would never do) - there are no experimental data on mechanosynthesis to analyze and there are no schemes in place to produce the experimental data!)

2. (Im)possibility . Let's be clear. Can I absolutely rule out the development of an MNT capability in a small subset of materials systems? No - I have never suggested otherwise. (Richard Jones has adopted a similar stance. See here ). Is Drexler's assertion that MNT will be able to construct virtually anything correct? No. How could it be? That statement violates key principles regarding dangling bond densities and surface physics laid out by Drexler himself in Nanosystems . Those principles were subsequently reiterated by Merkle in his rebuttal to Sci. American's "Waiting for Breakthroughs" issue in 1996.

Best wishes,

Philip

PS I was never quite certain what it was about the Moriarty surname that Conan Doyle disliked..!

Philip Moriarty said...

Howard,

Drat! I lapsed into tech-speak for a second. Taking your comments above into consideration, I'll clarify:

Dangling bond densities

Imagine taking a block of material (silicon, diamond, gold ..etc..) and cutting through it to expose a surface. The atoms at the surface have lost some of their neighbouring atoms (i.e. the surface atoms have 'dangling bonds') and are thus in a much less stable state than when in the bulk of the material. In the vast majority of cases (and there are many, many beautiful STM images that illustrate this) the surface atoms will rearrange (the technical term is 'reconstruct') to return to a lower energy (i.e. more stable) state.

Best wishes,

Philip

Philip Moriarty said...

Hi Damian,

Just checked out your website . Anyone who drums, has an interest in eighties metal, and cites Bon Scott is certainly somebody with whom I'm more than willing to continue nanotech-related discussions!!

(Re. drum solos: you're not into Rush or King Crimson, perchance?!)

Best wishes,

Philip

[P.S. Apologies for going off-topic, Howard, but 'arch-enemy' was certainly too harsh a term. (Unless Damian professes a liking for Motley Crue...)]

Howard Lovy said...

Well, I'll forgive it this once, Philip, but never, ever mention Rush again! Brings back too many high school memories. A buddy of mine was so into Rush, at random times he would squawk out those lyrics in that Geddy Lee falsetto. I tried to understand the lyrics. God knows I tried. I even read the liner notes for hidden Kabbalistic codes embedded within the meaningless sprawl of disjointed words. Nothing, nothing but the sounds of salesmen ...

Philip Moriarty said...

Across the River Styx, out of the lamplight
His nemesis is waiting at the gate
The Snow Dog, ermine glowing in the damp night
Coal-black eyes shimmering with hate


C'mon Howard, what's not to love about that particular gem?! And the drumming... no-one has the right to be that talented...

Nevertheless, I'm quite aware that Rush are... how shall I put it ... an acquired taste. I promise never to speak of them again (at least on Nanobot ).

Philip

Anonymous said...

Dear Howard,

My new main man Philip and I will discuss the matter further while waiting in line for the doors to open at our neighborhood arena, since it is obvious he is a man of exceptional taste, wit, style, class, and intellect (although, in this sad age of modern music and unappreciated talent, we'll actually be waiting at a table in a local bar). I used to know a guy that had a picture of Neil Peart (that's Rush's drummer, just in case) in his wallet. Right above the picture of his girlfriend.

Fell enriched, Howard. I had to listen to people doing bad Eddie Vedder and good Kurt Cobain impressions when I hit high school.

Aahhhhhhh SALESMEN!

Damian Allis