su3su2u1 is a pseudonymous Internet author who posts to many places, most notably Tumblr. He has argued, at great length, that MIRI is not a real research organization and that Eliezer Yudkowsky is a crackpot. Many have written responses, including me and Scott. Instead of writing yet more replies to su3su2u1’s claims about MIRI, I’d like to explain why everyone arguing with him should stop wasting their time.
EDIT: I should emphasize that the point of this post isn’t to criticize su3su2u1 just for the sake of it, but to save people from wasting their time arguing with him. Since this is my own advice, I will follow it, and not comment further on su3su2u1’s writings after today.
Although Eliezer is not a crackpot, I think everyone must admit that he (though not MIRI’s other researchers) has several apparent signs of crackpottery. These include not having a formal education; writing mostly for his own websites, instead of peer-reviewed journals; not having had an established reputation within AI, when he first started writing about AI safety; and not co-authoring his papers with well-known AI researchers. su3su2u1 frequently criticizes MIRI and Eliezer on these grounds. su3su2u1’s stated theory is that he criticizes MIRI and Eliezer because of these signs of crackpottery. Others have argued that these are merely excuses, and that su3su2u1 just criticizes because he dislikes Eliezer personally. If the “crackpot signs” all went away, under this alternative theory, su3su2u1 wouldn’t change his mind; he’d just make up new reasons for claiming Eliezer/MIRI are crackpots.
To test these theories, we could, if you will, imagine an alternate-universe Eliezer – an Eliezer-Prime – who has unconventional ideas, but none of the “crackpot signs”. For example, we could say that:
– While Eliezer himself was mostly self-educated, Eliezer-Prime got his degrees from MIT, one of the top technical universities in the world.
– When he first thought of “Friendly AI”, Eliezer wrote about it on his own website. But when Eliezer-Prime got his big idea, he instead first published it in the Proceedings of the National Academy of Sciences, one of the top few most respected scientific journals.
– su3su2u1 has said that, if he were running MIRI, his first priority would be to make everyone get a PhD. We can say that Eliezer-Prime got a PhD. For extra impressiveness, let’s say that he also got his PhD from MIT.
– Of course, a PhD might not mean that much, if it’s in an unrelated subject. If you’re an expert on X, you might still be a crackpot on Y. So we’ll specify that Eliezer-Prime was actually awarded his MIT PhD for his “crazy ideas”. His PhD thesis, of course, would have been approved by his doctoral committee, all MIT professors. We’ll pick the most renowned professors we can find for his committee, like Marvin Minsky and Gerald Sussman.
– Eliezer’s publications on Friendliness mostly aren’t peer-reviewed. But Eliezer-Prime’s are, of course. Eliezer-Prime has a track record of many relevant, technical, highly-cited and peer-reviewed publications, in respectable scientific journals.
– MIRI, which employs Eliezer, is relatively new and not that prestigious. Instead, we’ll give Eliezer-Prime a job at, say, Oxford, the oldest and arguably most prestigious university in the Anglosphere, with previous positions at other top universities like Stanford.
su3su2u1 might not agree with Eliezer-Prime, but hopefully, he wouldn’t just dismiss him as a crackpot. If he disagreed, he’d treat it like a serious discussion with a well-respected researcher in the field, and back up his points with technical, peer-reviewed sources.
Alas, it is not to be. Eliezer-Prime is real – his name is Dr. Eric Drexler, the founder of the field of nanotechnology – and su3su2u1 commented thus:
“The whole thing [molecular manufacturing] is pseudoscience. The founder, Drexler, is a crackpot himself.” [EDIT: su3su2u1 says that this quote is from a different person with the same username. However, he’s also said that he does in fact endorse this quote, so I am not misrepresenting him. See discussion below – my apologies for any confusion.]
In his dismissal of Drexler, su3su2u1 included no math, no equations, and no technical work. He based his arguments on loose verbal analogies, and (unsourced) claims that Drexler was unaware of even the basics of the field he invented, like scaling laws (which his MIT PhD thesis spends an entire chapter on). He cited only one source, an article by chemist Richard Smalley, which wasn’t technical and wasn’t even peer-reviewed. Rather, it was a two-page pop sci magazine piece, centered around a silly analogy comparing molecular chemistry to romance. Via Tumblr, I politely asked su3su2u1 for links to technical, peer-reviewed sources that rebut Drexler’s ideas; he has so far declined to reply. If he ever does, I’d be happy to post them below. But until then, it seems safe to say that the “making up excuses” theory is vindicated, and that trying to change su3su2u1’s mind just isn’t going to happen, no matter how many of his arguments are proven false.
EDIT: su3su2u1 has written a response, with two major points.
The first is that he continues to dismiss Drexler, and continues to not provide any technical arguments or peer-reviewed sources as a basis for his dismissal. He says that, “I believe you can find similar physicists making the same argument by walking into a material science department and asking any physicist about it.” But if that’s true, where are the peer-reviewed sources? Per Google Scholar, Drexler’s book Nanosystems (an edited version of his PhD thesis) has been cited over 1,700 times. His nontechnical book, Engines of Creation, has been cited over 2,200 times. His original 1981 paper in PNAS has been cited over 500 times. If su3su2u1’s opinions are common among physicists, surely there’s a peer-reviewed source which discusses them, somewhere in all those thousands of cites.
The second is that he says su3su2u1 is a common username, and one of the sources I cite is not actually him, but a different person using the same handle. [EDIT: This discussion has been moved to the comments, per Douglas Knight’s recommendation.]
(tumblr user somervta here)
I’m not sure you’re accurately representing su3su2u1 here. There’re two issues; firstly, the question of MIRI’s competence – how good they are at producing relevant research and achieving their other goals, and secondly, the question of the truth of MIRI-esque claims about AI. You’re criticizing him for not being consistent wrt his objections, but he’s objecting to Kurzweil on the latter ground, and most of his objections about academic respectability/crackpotness are about the former issue.
(also that person on the xkcd forums might not be the tumblr user su3su2u1 – xkcd!su3su2u1 was a regular reader of OB and LW, and IIRC tumblr user su3su2u1 was not.)
Sorry, I’m not sure I understand. By Kurzweil, do you mean Drexler? su3su2u1 objects to MIRI on both object-level and academic credibility grounds, and also objects to Drexler on both object-level and academic credibility grounds. The difference is that, on the object level, Drexler’s work is highly technical, while the arguments for AI safety being as important as MIRI thinks it is are largely non-technical and philosophical. Technical work is obviously easier to check for accuracy, so it’s much easier to see that su3su2u1’s criticisms of Drexler hold no real weight. And on the credibility level, since Drexler has a PhD from MIT, su3su2u1 is reduced to arguing that because his PhD was from the MIT Media Lab, it doesn’t count (yes really):
“But haven’t degrees been awarded for Drexler’s stuff? Aren’t there active researchers? Basically none of the active researchers have actual phds in a relevant subject. Drexler does have a phd but its from MITs Architecure and Planning college (specifically, the media lab. Anything crackpotty that comes out of MIT comes out of the medialab).” (cite: http://archives.somethingawful.com/showthread.php?threadid=3627012&userid=209704&perpage=40&pagenumber=1)
(Drexler’s full doctoral committee was Minsky, Sussman, https://en.wikipedia.org/wiki/Stephen_Benton, http://chemistry.mit.edu/people/danheiser-rick, and https://en.wikipedia.org/wiki/Alexander_Rich – hardly intellectual lightweights.)
su3su2u1 Tumblr has read posts from LW, and has commented there previously: http://su3su2u1.tumblr.com/post/128543655213/when-was-the-last-time-you-changed-your-mind-about
Alyssa, what is your actual, object-level opinion of Drexler? Also, how much chemistry did you take in college?
I think we might be able to do molecular manufacturing basically as Drexler describes it, but at this point we still don’t know, the technology just isn’t developed enough yet.
FWIW, I took some classes in chemistry and came in sixty-somethingith in the US Chemistry Olympiad, but am by no means a professional expert. If we’re arguing from authority, though, my personal authority isn’t really relevant here. Rick Danheiser (http://chemistry.mit.edu/people/danheiser-rick) is a famous MIT chemist who was on Drexler’s doctoral committee, and signed off on Drexler’s nanotech thesis, later published as Nanosystems. It’s entirely possible Drexler is wrong, but people who disagree with Drexler’s ideas should argue the point with Danheiser and all the other academics who’ve endorsed them (or at least not found basic gaping technical flaws in them), not with me.
Although I agree with some of your points, I’m not sure your argument form is valid. It sounds like you’re saying that if there are any credentialled establishment people you think are crazy, you can’t also believe that the typical signs of crackpottery have any predictive power. The obvious counterargument is that some establishment people are crazy, but many more crackpot-model-fitting people are.
Also, even if you were right I think it’s a big leap from “there is a tension in his position on these two issues” to “don’t bother arguing with him, he’s too biased”. This is especially problematic because it seems reminiscent of su3su2u1’s own argument style – catch Eliezer or some LWer in what he thinks is an error, then use it as a blanket dismissal of everything he’s ever said. Maybe you’re intentionally parodying that, but if so I think it’s too subtle for me.
It’s not just “any credentialed establishment people” – if su3su2u1 thought that, eg., a random physics professor was a crackpot because of his ideas about string theory, the argument would be less strong. Drexler is in MIRI’s social sphere (I’ve personally met him and he works for FHI), and like MIRI, he is weird in the specific sense of having ideas about very powerful future technologies that sound crazy to a naive observer, due to being many decades removed from our current technology base. If he believes both MIRI and Drexler are crackpots, this belief is much more likely to be explained by a shared trait of MIRI and Drexler than if he thought both MIRI and Physics Prof. Doe were crackpots.
I don’t think this means people should dismiss everything su3su2u1 has said ever. Eg., if he has any physics writing, it wouldn’t be fair to dismiss it on these grounds. But I never claimed that – I just claimed that people shouldn’t bother arguing with him about MIRI. His criticisms of MIRI and Eliezer are *very* lengthy (eg. his chapter-by-chapter “review” of HPMOR), and probably do make up a large fraction of his writing.
You are underestimating the number of crackpot-like individuals that haunt universities, often in prestigious positions where they have accomplished great work earlier in their lives.
For example, Don Anderson, at the end of his career, was completely, stubbornly, hopelessly wrong (about mantle plumes), to the point of antagonizing and belittling colleagues at conferences. Yet he had a incredibly brilliant career and contributed enormously to the field. He’s far from the only one.
You can easily find the texts of the Drexler-Smalley debates. They are quite poor on the maths side, as if both are thinking that the arguments given by the other do not reach a level where this is even necessary. Depressing stuff.
Re crackpots, see my reply to slatestarcodex. su3su2u1 doesn’t just think Drexler is wrong. He dismisses him as a crackpot who makes basic technical mistakes, but can’t provide any peer-reviewed sources that point out these mistakes, out of the thousands of people who’ve cited Drexler’s work.
I’ve read Smalley’s writing on Drexler, and as I said in the post, it doesn’t make detailed technical arguments and isn’t peer-reviewed. Drexler has spent four decades making technical arguments in top scientific journals. If he’s wrong, he deserves a response with at least one out of the two.
I think when su3su2u1 mentioned the Smalley/Drexler debate he meant this
https://en.wikipedia.org/wiki/Drexler%E2%80%93Smalley_debate_on_molecular_nanotechnology
FYI, su3su2u1 on Tumblr is not the same as su3su2u1 on the xkcd forums.
Replied via an edit to the original post.
Accusing people of this kind of deceit is not worth it. It just damages your reputation, even if it is true. I think that a coincidence here is absurd, but there are other possibilities, like impersonation. Move the paragraphs of quotes into the comments, or something.
[Hi Douglas – WordPress won’t let me reply for some reason, so I’m replying via edit. Thanks for this note; I’ve moved those quotes into a comment per your suggestion.]
su3su2u1 has said he IS the SomethingAwful user, but is not the xkcd user when asked on his tumblr. I don’t know what he has to gain by owning up to one but not the other.
@FrobeniusMethod: Thanks, but can you please cite sources? Sources are important here.
Tumblr doesn’t have a search feature, but if you dig far enough back, it has come up before.
For a recent link on his tumblr
http://su3su2u1.tumblr.com/post/128668939353/yeah-she-thinks-youre-just-lying-about-not-being
su3su2u1 has made many arguments about many LW-related topics, many of them more technical and inside-view than the ones you refer to here (for instance, the numerous criticisms he makes in his HPMOR liveblog). Unless you have a more general theory that explains why this whole miscellany of criticism is all simultaneously not worth engaging with, isn’t it surely too strong to say “don’t bother arguing with him”?
(In other words, I agree with Scott’s second point — “this person has made a bad argument” does not imply “no one should bother arguing with this person,” even if the premise is true.)
I think I’ve addressed these points, in my replies to Tarn and slatestarcodex
FWIW, I wrote a longer reply to this post here:
I apologize for the intemperate tone. I was kind of worked up when I wrote it.
It also could be that personal hostility motivates him to search for the best reasons that can be found and use them. It may then be that some targets of his personal hostility have more good reasons to be found than other targets.
In other words, it might be that his criticisms of Drexler aren’t very good, but his criticisms of Eliezer *are*–because Eliezer happens to be a target that is more worthy of criticism and for which it therefore is easier to find good criticism. Even if the criticism is not his true rejection, you’re not just arguing with him; you’re implicitly arguing with the members of the audience who support such criticism, and for those members, the criticism really *could* be their true rejection.
Yeah, su3su2u1’s criticism of drexler is not really his criticism, pretty much everyone in nanotech/materials science feels that way. Nanotech is a massive and highly profitable research field in which virtually no one works on drexler-style “molecular assemblers,” because they are widely viewed to be impossible.
I explicitly said that disagreeing with Drexler wasn’t the problem; I wouldn’t be that surprised myself if Drexler turned out to be wrong. The problem was dismissing him as a crackpot who makes fundamental basic mistakes – mistakes which were somehow not caught by the famous MIT physics professor, famous MIT chemistry professor, and famous MIT biology professor who all signed off on his thesis – and not providing any technical arguments or peer-reviewed sources to back up his dismissal. If so many people agree with this criticism, why has no one provided any peer-reviewed sources, out of the thousands of people who’ve cited Drexler’s work?
This is a pretty polemical piece. Smalley is qualified, he’s not just some random chemist with a pencil, he won a freaking Nobel Prize. Smalley’s argument is that chemistry is very difficult and imprecise, Drexler’s counterargument is that precise nanotech is difficult but not impossible. Whether you expect AI to build nanotech and then take over the world depends on how smart you expect it will be, which is probably the main underlying disagreement present here.
I never said Smalley wasn’t qualified. I said that Smalley’s criticism of Drexler was a two-page article in a pop sci magazine, it didn’t go into technical detail (to address Drexler’s 500 pages of detailed, technical argumentation), and it wasn’t peer-reviewed, all of which are true. If Smalley has any technical, peer-reviewed work that criticizes Drexler, I’ll happily edit the post to link to it, but I’m not aware of any such and the Wikipedia article on the Drexler-Smalley debate doesn’t cite any.
Smalley’s “piece” is a debate carried on with Drexler over several years. It is not a two page piece. The debate has it’s own wikipedia article. It carried across several publications.
su3su2u1 is walking the the physics party line with the objections that physicists always raise about atomically precise molecular manufacturing, as Phil said above. I don’t understand how a physicist saying the things physicists always say about nanotech disqualifies his views on MIRI.
I’ve noticed a troubling trend in the rationalist community toward ad hominem fallacies. Sometimes it takes the form of a signalling argument, and sometimes it’s attacking motivations, but the end result has consistently been to claim that we don’t need to listen to criticism because of the source. This one is particularly troubling because it doesn’t even seem relevant.
It’s difficult to come up with specific criticisms for non-existent nanotech blueprints. The burden of proof is on Drexler’s camp to prove that precise nanotech will be possible, given all the difficulties associated with it, and so far that has not been done. Note that Smalley does think nanotechnology is a good idea for some applications, he just thinks it won’t be either a panacea or doomsday device. It’s possible that with a supergenius AI the difficulties can all be overcome and we’ll be able to use nano replicators to do any arbitrary task we want, but I think it’s reasonable to find that possibility improbable.
I am favorably following the LW sphere for quite a while, but you seem to intentionally misquote his tumblr posts in question here.
“su3su2u1 has said that, if he were running MIRI, his first priority would be to make everyone get a PhD. ”
When I read this post, he is saying the following:
He thinks MIRI would be more effective if they would be connected to the academic research world of AI and machine learning in the sense that people working for MIRI should have the domain knowledge of current research in those fields as well as personal contacts. While MIRI, their work and their papers should be well known within those communities. (A good argument why this statement is false would be interesting)
He states that this isn’t the current state at all (which I cannot judge) and that the easiest way to get to there, is by sending MIRI employers as PhD students to excellent research labs. Because the everyday work of a PhD student is exactly the exchange with academic research. So even if they wouldn’t finish their PhD at all, this would still fulfill the purpose.
An far stretched analogous would be,
He says: you would learn a lot about practical machine learning by participating in kaggle challenges.
You say: he looks me on them because I haven’t been a nominal member of a winning team at kaggle, he only cares about credentials!!!
Hello. Sorry for failing to cite, but I think I was referring to a different post (link: http://su3su2u1.tumblr.com/post/128170209213/by-some-twist-of-fate-you-wake-up-tomorrow-and), where su3su2u1 said explicitly:
“We can fix that with one approach- every researcher [at MIRI] without a phd is going to school. Since they’ll be supported by MIRI and MIRI has some ties to professors at institutions it shouldn’t be so hard to get people enrolled. I’d probably give them 4 or 5 years to finish a phd in CS/AI (they won’t be teaching because they have the outside support, so they just need to pass the core classes, publish papers/get into some conferences, graduate) or the like.”
This was the very first recommendation he made when someone asked “By some twist of fate, you wake up tomorrow and find you’ve been appointed director of MIRI. How do you run things?”. I think saying that “if he were running MIRI, his first priority would be to make everyone get a PhD” is a reasonable and fair interpretation.
Read the text above that quotation:
“Anyway, the big problems I see are that there is not enough experience at actually accomplishing research in the organization, and they seem to insist that they can somehow do impactful research largely on their own without stronger ties to the academic community. ”
Now read the text below that quotation:
“That will
1. dramatically boost publications with the MIRI imprint
2. Get more ties to academia and more coauthors on MIRI papers
3. give MIRI some results in reasonable publications ”
kingnothing seems to give a better characterization than you.
The second point that su3su2u1 makes in his reply is that he says su3su2u1 is a common username, and one of the sources I cite is not actually him, but a different person using the same handle. Here, I must admit confusion. If this is true, and su3su2u1 points out the specific source, I will of course remove it and apologize. However, the three sources I cited shared not just a username, but similar-sounding criticism of MIRI and Eliezer (and Drexler), who are obscure enough that average physics graduates wouldn’t know about them (“su3su2u1” is a particle physics reference). It seems unlikely that two people would share both the same particle physics username and the same criticism of obscure AI researchers by pure chance, but I guess anything is possible.
EDIT: Per a comment on this post by FrobeniusMethod, su3su2u1!Tumblr has admitted to being the same person as su3su2u1!SomethingAwful, but continues to deny being the same person as su3su2u1!XKCD. Although the criticisms of Eliezer, Drexler, and Bayesian statistics that su3su2u1!XKCD makes are extremely similar-sounding to su3su2u1!Tumblr, su3su2u1!Tumblr has said that he endorses the relevant quotes by su3su2u1!XKCD, and so the question of identity is largely moot because su3su2u1!Tumblr has confirmed that opinions aren’t being misrepresented. I’ve removed the links to su3su2u1!XKCD to prevent any possible damage from mistaken identity.
The first source is his main blog on Tumblr: “Based on their output over the last decade, MIRI is primarily a fanfic and blog-post producing organization. That seems like spending money on personal entertainment.” (cite); “What I was specifically thinking was “MIRI has produced a much larger volume of well-received fan fiction and blog posts than research.” That was what I inended to communicate, if somewhat snarkily. MIRI bills itself as a research institute, so I judge them on their produced research. The accountability measure of a research institute is academic citations.” (cite); “It’s pragmatic, use whichever model of reasoning under uncertainty you like but check to make sure it’s doing what you want. If Bayes fails the check, chuck it out. I claim that this is strictly superior to philosophical Bayesianism because it provides real guidance for working on problems, unlike Bayesianism which gives you an impossible to follow recipe “start by finding this uncomputable metric for every single possible hypothesis.” I can make the stronger claim that pretty much everyone involved in serious exploration of the world seems to be following my method.” (cite); “BUT- nanotech, in Drexler’s vision is a silly idea. Drexler wants what he calls “dry” nanotechnology (…) It is fundamentally silly to think that we can apply systems and principles relevant at the meter scale down to the nanoscale. (…) How does nature do it? Nature does what Drexler called “wet” nanotechnology. It is nothing like mechanical engineering principles (and pointing to it as an example of nanotech is very misleading). Everything is powered by the solvent water.” (cite)
The second source is the XKCD forums: [EDIT: These links have been removed, to prevent any possible damage caused by mistaken identity. See edit at the top of this comment.]
The third source is the SomethingAwful forums: “[Eliezer Yudkowsky] doesn’t know enough to write sophisticated code. He wants to get paid for blogging and for writing Harry Potter fanfiction. Look at his ‘revealed preferences’ to use the economic parlance. When given a huge research budget, instead of hiring experts to work with him and doing research, he instead blogs and writes Harry Potter fanfiction. The people [MIRI] hires are for the most part undergraduates with bachelor degrees, not domain experts, not proven researchers. The requirement to get hired seems to be long-time involvement in his “rationalist” community. This is why even the paper [MIRI] has submitted to peer review is poorly written- no one at the institute knows how to first author a paper. (…) I’m willing to bet both of us have published more peer reviewed papers than Yudkoswky’s entire institute.”; “Sure, a lot of Bayesian ideas are incredibly useful when you have the right application. The key is to have the right application. I make a lot of statistical models for work (like many physicists I work for an insurance company predicting high risk claims), and sometimes Bayesian modeling really is the best way to go. Those situations are rare though, and if I insisted on Bayes as the one true probability model, I’d be fired.”; “Drexler’s idea is that we can take the mechanical engineering manufacturing paradigm and scale it down to the very tiny. (…) The important point I want to make is that things are very different at small scale. (…) BUT, the LessWrongian might ask- what about biology? Good question- biology operates completely differently. Everything is wet, and the unique properties of water are an integral part of how everything works. (…) “Real” nano-tech will look more like the chemistry of enzymes in water. (…) But haven’t degrees been awarded for Drexler’s stuff? Aren’t there active researchers? Basically none of the active researchers have actual phds in a relevant subject. Drexler does have a phd but its from MITs Architecure and Planning college (specifically, the media lab. Anything crackpotty that comes out of MIT comes out of the medialab). Like all things transhumanist, the entire field is basically a (perhaps unintentional) con game.” (cite)
Background: I am qualified to peer review nanotech publications and have done so many times. I am a materials scientist who wears a lab-coat almost every day.
I dislike relying on peer review as an unbreakable backstop. Yes, it’s an important step in vetting material, but it isn’t the end-all. As you know, the fact that a belief made it through peer review doesn’t make it true. It also doesn’t reflect a general scientific consensus, which is the actual thing you should most often trust. I can confirm su3su2u1’s statement about actual lab-coat wearing materials scientists and what they think of Drexler’s nanomachine stuff. His and my viewpoint on this is not unanimous, but widely held. So why doesn’t it appear in the peer reviewed literature? From my perspective: why the hell would anyone go to the trouble? Most materials scientists are too busy doing actual science to spend time publishing untestable opinions about the future of nanotech. Smalley took the time to do this (and he was an uncharacteristically philosophical materials scientist) in the C&EN column. Beyond that there isn’t much reason to do so. It won’t advance your career. It’s not worth arguing over such unlikely hypotheticals. So you don’t find these refutations in the literature. Most scientists consider it a matter of opinion (and frankly, it’s not bad for your field to take on a mythical status, true or not. “Nano” has gotten sooo much money just for being nano.).
So a question for you: what is more valuable in shaping your opinion, the fact that no materials scientists have taken time to publish their personal response to Drexler, or the fact that most of them are highly skeptical of his point of view? Or if this is new information to you, do you consider it relevant to your opinion?
(Note: I’ve verified that AlkalineHume does have real science credentials.)
Thanks for your comment – I think it’s good to hear from someone with real experience in a relevant field. Of course, I agree that wrong ideas sometimes get through peer review. For that matter, as I’ve said in earlier comments, I wouldn’t be all that surprised myself if Drexler turned out to be wrong about the practicality of nanoassemblers. But the problem isn’t that su3su2u1 thinks Drexler is wrong – scientists think other scientists are wrong all the time. The problem is that su3su2u1 dismisses Drexler as a “crackpot” who isn’t even worth taking seriously, in the same way one might dismiss Gene Ray’s Time Cube, despite Drexler having every conceivable qualification in the field he largely invented. The problem is that su3su2u1 claims to have identified basic, fundamental science errors in Drexler’s work, without even directly citing any specific paper by Drexler, and without any explanation as to how these fundamental errors weren’t caught by the six separate MIT professors who reviewed Drexler’s thesis, any of the peer reviewers who looked at Drexler’s dozens of publications, and any of the thousands of academics who have cited Drexler.
I’ve heard the “everyone is too busy” theory before, but I’m frankly skeptical of it. As I noted in my edit to the main post, Drexler’s book Nanosystems (an edited version of his PhD thesis) has been cited over 1,700 times per Google Scholar. His nontechnical book, Engines of Creation, has been cited over 2,200 times. His original 1981 paper in PNAS has been cited over 500 times. Clearly, many, many people have seen his work and think it’s worth discussing. And, quite frankly, it seems pretty likely that many scientists now working in nano wouldn’t be there if it weren’t for funding caused by Drexler’s advocacy during the 90s. I feel like a standard which lets Drexler be dismissed, on the grounds that “it’s clearly wrong, but nobody has the time to write a paper refuting it”, is so broad that it could be used to dismiss almost any idea, even foundational ones like the Big Bang theory. How can science ever arrive at the truth, if everyone simply dismisses any new idea, no matter how well-argued and well-supported, without bothering to write a real refutation?
And yes, I feel that, if someone has gone through the trouble of writing up detailed, technical literature that makes it through peer review, it’s academia’s responsibility to reply in kind, before simply dismissing the author as a “crackpot” who isn’t even worth talking to. Daryl Bem’s work on ESP is, by any standard, far more radical than anything Drexler has written. If Bem is correct, we need to throw out several fundamental laws of physics (and for that matter biology), like the unidirectional flow of time. But Bem took the time to do his experiment, and he wrote it up with statistics and technical details in place of anecdotes, and he got it through the official peer review process. And so Wagenmakers looked at his claims, and wrote a detailed, thorough, technical refutation, which he also had peer-reviewed (http://www.ncbi.nlm.nih.gov/pubmed/21280965). And I think the world is much better off for him having done so – even though I think Bem is obviously wrong, Wagenmakers’ paper pointed out several major changes that we ought to make to psychological research generally.
“But the problem isn’t that su3su2u1 thinks Drexler is wrong – scientists think other scientists are wrong all the time. The problem is that su3su2u1 dismisses Drexler as a “crackpot” who isn’t even worth taking seriously, in the same way one might dismiss Gene Ray’s Time Cube, despite Drexler having every conceivable qualification in the field he largely invented.”
In his debate with Drexler, Smalley accuses Drexler of being “in a pretend world” and selling “a bedtime story.” I might be misreading this, but that sounds like Smalley politely telling Drexler “you’re nuts,” as directly as possible given the venue.
On top of that, Julius Rebek, who was part of MIT’s chemistry department at the time Drexler got his Ph.D., has been quoted in Wired magazine saying that Drexler’s thesis, “showed utter contempt for chemistry. And the mechanosynthesis stuff I saw in that thesis might as well have been written by somebody on controlled substances.”
So su3su2u1’s assessment of Drexler doesn’t sound like it’s outside the range of things respectable scientists have said.
Thanks for your reply. I can appreciate that it’s difficult to swallow that no materials scientists are down for this debate. Perhaps it is a little strange, but I think that’s partly due to a disconnect between the general materials science perception of Drexler and the more lay-person perception. The general materials science perception of Drexler is, and I quote, “who?” His nano-related work is mostly non-technical and sort of meta-sciency. When I think of the “founders” of nano, Drexler doesn’t crack the list by a long shot. Besides the Feynman talk, you have Smalley, Louis Brus, Paul Alivisatos, Moungi Bawendi. They all have boatloads of students who’ve gone on to make big contributions. (My list is very chemist-biased, so if there are any top-down nano people in the crowd I apologize.) But Drexler isn’t a scientist, frankly. Not that that’s a condemnation, just that it’s hard to found a science field without doing much science. That’s why the “everyone is too busy” argument holds up. Yes, they’re too busy to take on a debate with someone in a totally different non-science field that honestly isn’t all that relevant to what they do. (And note that to debate that point you have to argue that most materials scientists don’t agree with my position, which I assure you they do.)
So you might ask, if they’re too busy to refute him why are they citing him? I’d be curious to know how many of those citations are from actual science papers as opposed to meta-sciency stuff. They could all be, I really don’t know. But for those that are hard-science citations, it comes back to incentive. If you’re writing an intro to a paper you’re going to cite people who support you. Drexler clearly thinks nano is a big deal, so it might make sense to cite him if you want to support that viewpoint and if you like his work. A couple thousand citations (modulo however many are meta-science) is honestly not all that many for a “major” nano paper from decades ago. That’s a hundred a year. There are thouuuusands of nano papers each year. Nano Letters alone must publish 30 a week. So yes, he’s being cited, but not all that much and it’s not clear by whom.
So let’s get back to the su3su2u1 criticism. I haven’t read Drexler’s book, so I can’t really comment from a place of knowledge. But his criticism certainly resonates with me. There are absolutely interesting questions about what can be achieved. According to su* Drexler doesn’t approach from that angle. So his non-technical criticism is of an (apparently) non-technical book (again, haven’t read it). I don’t see much wrong with that. I think the main thing to keep in mind here is that Drexler not correctly categorized as a scientist. I don’t know that his ideas merit a highly technical criticism. They are just so far outside of what we can actually achieve technically that there’s not honestly all that much to say about them. I think this is where su*’s Drexler criticism and his MIRI criticism overlap. Both are responses to the development of highly technical fields that extrapolate along a very particular direction that may or may not pan out as the field develops. I have to say I side completely with su* when he says that the best way to figure out how to keep AI/nano/whatever friendly is to engage with those fields where they are, not where you think they might be in X decades. Because there are so many ways either field could go that it’s like trying to talk about what you should do in move 50 of a chess game that’s currently at move 20. For what it’s worth, I think much more important than the sticky fingers problem is a problem of symmetry. It is incredibly hard to make free-standing low-symmetry nano-objects out of anything but biomolecules. There are no really good ideas around this. If you can’t get low symmetry you really can’t get anything that looks like a tool. This is a big, big problem.
Presumably, there is some reason that biomolecules might not work well for nanotechnological applications. Would you mind explaining it?
I feel very uncomfortable about posts like this. Declaring that people who criticize a group aren’t worth arguing with is an easy way to create echo chambers, and this is not the first time I’ve seen this sort of thing being stated about a clearly educated critic of MIRI and Eliezer. In that regard, I found Scott’s blog post in contrast to be substantially more productive a framework. (Copying this comment over from Facebook to here at your request.)
Thanks Joshua. Copying over my reply from Facebook:
In general I’m inclined to agree, but in this case, su3su2u1’s criticism of MIRI and Eliezer is so voluminous that pretty much every point has already been made (and most have already been responded to). If he had just written one or two posts criticizing MIRI, I would outright agree with you. But su3su2u1 has, eg., in addition to all his direct criticism of MIRI’s work, written an entire chapter-by-chapter hate review of Eliezer’s fan fiction HPMOR, which has 122 chapters and is longer than the entire Lord of the Rings, the Silmarillion, and the Hobbit combined. At some point, you just have to give up and move on to other concerns.
There have been cases like this before; eg. Richard Loosemore was a longtime poster to Eliezer’s SL4 list a decade ago. He was definitely educated, and claimed to be an AI researcher, and kept criticizing MIRI and Eliezer in the same name-calling, dismissive manner. After many, many months, and many, many discussion threads, he was eventually banned. It turned out that his “AI research” contained no technical content at all, and was basically all him yelling at people he didn’t like in PDF form (feel free to verify this via Google). He has continued writing “papers” that are basically just yelling and name-calling at MIRI, but they are now mostly ignored, and it seems clear that everyone is better off for it.
Eh, it sounds almost like you’re saying that folks shouldn’t argue with su3su2u1 about MIRI because HPMOR is very long. But surely su3su2u1’s quip about HPMOR being MIRI’s most central project is just a mean joke, not The Truth About MIRI. So, what’s wrong with ignoring the fanfic reviews and engaging only with relevant criticism?
Not 100% relevant: I realized today that prions or prion inspired methods might be well suited for overcoming some of the difficulties nanotechnology faces. Prions are well-organized replicators, highly resilient, yet very small. It seems like a possible bridge of sorts between wetware and hardware. I did a couple cursory Google searches, and it turns out that Susan Lindquist, a professor of biology at MIT, is currently working in the lab on using yeast prions to build an actual circuit. That’s pretty neat.
Really liked the way how you took a claim and tore it apart with one example.