Want to read Slashdot from your mobile device? Point it at m.slashdot.org and keep reading!

 


Forgot your password?
Close
typodupeerror
×
Science

Wolframania 255

Posted by michael from the making-a-splash dept.
An Anonymous Coward writes "The New York Times has had a couple of articles about Stephen Wolfram in the last couple of weeks. Is he self-aggrandizing or brilliant? Or both? And is God a software engineer?" I thought our reader-contributed review of ANKOS was quite good.
This discussion has been archived. No new comments can be posted.

Wolframania More

Wolframania

Comments Filter:

  • Praise, either way... (Score:3, Flamebait)

    by killthiskid ( 197397 ) on Sunday June 16, 2002 @10:10AM (#3710863) Homepage Journal

    A quote from each article:


    Had Dr. Wolfram been more demonstrative in parceling out credit to those who share his vision (many are mentioned, in passing, in the book's copious notes), they might be lining up to provide testimonials. It's the kind of book some may wish they had written.

    Yet Wolfram has earned some bragging rights. No one has contributed more seminally to this new way of thinking about the world. Certainly no one has worked so hard to produce such a beautiful book. It's too bad that more science isn't delivered this way.

    Everywhere you look, almost everyone is saying, well, even if he is wrong, he's written a hell of a book. Which I suppose is true.



    • For peer review ... (Score:1, Interesting)

      by Anonymous Coward
      Wolfram's book is a very long slog. For inspiration to keep the appetite whetted have a look at Barabasi's - Linked: The New Science of Networks (http://www.nd.edu/~networks/linked/). The nice thing is you can also look at some of his papers in peer reviewed journals at http://citeseer.nj.nec.com/cs where you can see that he has been cited in the works of others. It may give you a better feel for the merit of Wolfram's tome by comparing his work to that of a peer-reviewed colleague.
    • Everywhere you look, almost everyone is saying, well, even if he is wrong, he's written a hell of a book. Which I suppose is true.

      1250 pages represents an awful lot of wasted time if he's not right. From what I've read, it seems Wolfram never sought peer review. That seems very curious to me. I think I'll wait for more reviews.
      • (day1) first 350 pages are a bit slow
        (day2) starts to hot-up around pg. 525
        (day3) pg. ~650, first application of this new learning
        ** there were a lot fewer pictures after this so now it's bathroom reading. i'm good 'til 2003 with this one.
    • It's too bad that more science isn't delivered this way.

      I must disagree with this statement. The way science is forged is by having other scientists spotting gaps in your arguments. That the gaps in an unreviewed work do not walk up to the non-specialist reader and introduce themselves does not mean they do not exist.

      • What peer reviewed journal would have published this 10 year 1000+ page opus? Who could fit it?

        • Re:Praise, either way... (Score:2, Informative)

          by ariels ( 6608 )
          The classification of the sporadic finite simple groups was published in peer-reviewed journals. I believe it is estimated to be around 10,000 pages. Nobody reviewed the whole thing in one fell swoop, of course.


          But each portion was reviewed. For instance, Walter Feit and John Thompson proved a first step as Solvability of Groups of Odd Order, Pacific Journal of Mathematics 13 (1963), 775-1029. I do not know if (or what) problems were raised regarding publication of such a long paper.

    • Re:Praise, either way... (Score:5, Interesting)

      by fatphil ( 181876 ) on Sunday June 16, 2002 @11:10AM (#3711044) Homepage
      "It's too bad that more science isn't delivered this way."

      Written by S. Wolfram,
      Peer reviewed by noone,
      Edited by S. Wolfram,
      Published by S. Wolfram's company.

      That's not the best route for 'science' to take in its delivery.

      Having said that, I think one comment that seems to be applicable is the ancient "both new and interesting; that which is interesting isn't new, and that which is new isn't interesting" style quote. He's very bad ad giving credit to those who did so much before he was even in nappies.

      FP.

      • He deserves praise either way despite his arroganc (Score:4, Interesting)

        by FreeUser ( 11483 ) on Sunday June 16, 2002 @08:16PM (#3712783)
        Written by S. Wolfram,
        Peer reviewed by noone,
        Edited by S. Wolfram,
        Published by S. Wolfram's company.

        Well, Wolfram had a team of PhDs working under him, so it did go through some nomimal review and quite a rigorous check for accuracy. That is certainly comparable to the "peer review" that one sees in publishing scientific papers in scientific journals, and is arguably better than much of the "peer review" that takes place prior to such publications.

        The real "peer review" will be that of other scientists now that his work is published. Can they replicate his results (almost certainly) and do the applications he outlined produce useful results to those working in the various fields of scientific inquiry his book touches upon. Quite possibly ... we'll just have to wait and see.

        I'm reading his book now, and it is quite fascinating. I disagree with the various calls for editors others have been making ... he is trying to drive a point home, and (thus far, I've only made it through chapter 7) is doing so in a time honored, rigorous fashion that is reminiscent of just about every theoretical mathematics, physics and engineering course I've taken.

        Does that mean his conclusions are correct? No.

        But it does set a very solid foundation for his thesis, and allow one to regard his theories in a solid context and an informed way, and, what is more, to understand them without first having become an expert in the field of CA.

        He thinks he's discovered an overlooked tool for doing scientific analasys of systems which to date have defied calculus and other methods of analysis. He makes a compelling argument for why this is so, and provides ample data and information for anyone who is interested to duplicate and check his work.

        He may not be correct, and his method of publishing may not have been within the channels the establishment generally prefers, but his publication itself appears to be in no way lacking in scientific rigorousness, and has certainly provided the detail and wherewithall for anyone to challenge it.

        He may not be paying proper homage to those who came before him, or giving sufficient credit to those who have thought along similar lines (though he does cite other works and give due credit, so I'm not sure that criticism is even accurate), but his work, right or wrong, certainly appears scientifically valid. And if it is wrong, it will be rebutted quite thoroughly I'm sure, given the number of toes he has likely stepped on in persuing such a nontraditional course.

  • pure liquid evil (Score:1, Funny)

    by Anonymous Coward
    Is he self-aggrandizing or brilliant?

    ... Based on the number of frosh at my alma matter that cringe at the mention of the name "Mathematica", I'd say he's the antichrist.

    "Life is hard, but it's harder if you're stupid"

  • SW's 256 autometa (Score:5, Interesting)

    by ajs ( 35943 ) <ajs@ajs . c om> on Sunday June 16, 2002 @10:16AM (#3710886) Homepage Journal
    He uses a classification of 256 particular 2D autometa for a lot of the examples in the book that's kind of interesting. I took the time to write some code for it [ajs.com] to explore the various permutations. It's CGI-based and it generates a png or jpeg image, so just throw it in your cgi-bin and check it out. The comments list the various options you can send it.

  • His Website.. (Score:3, Informative)

    by routerwhore ( 552333 ) on Sunday June 16, 2002 @10:17AM (#3710887) Homepage
    More insight on his website:
    www.stephenwolfram.com [stephenwolfram.com]

    Another good article about his latest work: On Forbes [forbes.com]

  • NyTimes Slashbox (Score:2, Offtopic)

    by asv108 ( 141455 )
    Looking over the front page, three out of the last eight stories are from the New York Times [nytimes.com]. Shouldn't slashdot just start recommending that geeks read the nytimes everyday? Every morning I read the paper and I can always pick out the stories that will end up on /.. We might as well go over some of the other cool Nytimes articles not mentioned yet on slashdot:

    • We might as well go over some of the other cool Nytimes articles not mentioned yet on slashdot: ... Review of a new book about the rise of eBay [nytimes.com]
      Actually, /. covered that with this story [slashdot.org] on Friday, reviewing the actual book that the NYTimes piece is based on.

      Maybe you should spend less time on NYTimes and more on /.

      :)

      Slashdot; who needs other sites?


    • How about an "NYTimes Registration-Free Mirrors" Slashbox?

  • science.slashdot.org (Score:1, Offtopic)

    by ajs ( 35943 )
    The front page has a link to this article through science.slashdot.org. That name does not resolve. The link is fine if you take out the "science". Hmm... I didn't actually mean that as commentary :-)
    • Yes, that's just it. I've read here and there in ANKOS and its absolutely fascinating, but is it science, or mathematics?

      If we define science in terms of observation and experiment, leading to theory, and then back to observation, does the "behavior" of a machine deserve to be included?

      If the book were titled "A new kind of mathematics (with scientific implications)" perhaps that would be more accurate?

      Where exactly is the science in ANKOS?

      Of course if it is really a NEW kind of science, perhaps we don't need observation of "real world" phenomena. But I'm troubled by that meaning of science.
  • Explaining the universe with equations is flawed; we should be explaining it a cellular automata, or as a computer program...

    but we know that cellular automata and computer programs can be expressed fundamentally as equations..... no?
    • I suppose Wolfram, being a physicist and therefore having no real clue of math, means that a model consisting just of smooth operator equations won't be sufficient.
      This theory shouldn't be rejected per se, because there could be very well some non-continuous at least at quantum level.
      But Wolfram obviously hasn't grasped the usefulness of a clean mathematical formalism. Otherwise his book won't be so diffuse.
      However Wolfram doesn't seem to understand the complexity which arises even from continuous systems and that in fact non-continuous dependencies can turn up in continuous systems. Do I even have to mention the Lorentz system at all, everyone should know it. But he is just a physicist after all.
      • being a physicist and therefore having no real clue of math

        some theoretical-physics profs I know actually have VERY deep understanding of math, which they use in their work.

        get to know the real masters, if you're smart as that mensa card implies, you'll understand how wrong was that remark.
      • Blockquoth the poster:

        being a physicist and therefore having no real clue of math

        That's sort of like saying, being a painter and having no real clue of paint. Archimedes? Newton? Maxwell? Laplace? Legendre? Einstein? It is no accident that major fields in mathematics have been opened up by ... wait for it ... physicists. It might be argued that physicists and their little problems have done more for the advancement of pure mathematics than all the scribblings of pure mathematicians. At the very least, an out-of-hand dismissal of physicists as, apparently, math-illiterates, is without justification.
        • Maybe he thought (with a good deal of justification) that physicists occasionally get a little too full of themselves, and need to be reminded that a lot of their work rests on the mathematicians who came before them.

          Einstein was a physicist, not a mathematician. He faced some limitations because of this. He wasn't able to do a lot of his most important work until he hooked up with real mathematicians such as Marcel Grossman, simply because he didn't have the depth of mathematics that they did.

          Let's look at your analogy; like a painter having no real clue of paint. How many of them actually do know how to actually make paint? Where would the Renaissance artists be without the guys who invented oil paints? The painters, like the physicists, only needed to know those properties which would affect their work.

          It might be argued that physicists and their little problems have done more for the advancement of pure mathematics than all the scribblings of pure mathematicians.

          Ah, "scribblings". There's an easy way to dismiss mathematicians as unimportant, without even having to back it up.

          At the very least, an out-of-hand dismissal of physicists as, apparently, math-illiterates, is without justification.

          Like you just dismissed mathematicians?
          • >How many of them actually do know how to
            >actually make paint? Where would the Renaissance
            >artists be without the guys who invented oil
            >paints?

            ah, there's nothing like a little uneducated blather.

            Most artists made their own paints. Many of them still do today.

            The guy credited with inventing modern oil paint was Jan van Eyck, an artist. Others were using oil paints before him, but he's widely credited with developing a stable oil-based varnish for use in them.

            There were later improvements by others, such as the addition of lead oxide by Antonello da Messina, da Vinci's addition of beeswax, and Rubens's grinding techniques. All of those guys, in case you're completely clueless, were artists.

            Think, then talk. It works better that way.

            Kuroth

        • Hang on, hang on....

          Newton: yes, he did a lot of maths, but (for instance) his notation for calculus was so clumsy that everyone ended up using Leibniz' notation.

          Einstein: he came up with the concepts, everyone else did the maths. Surely you know the old cliche of 'even Einstein flunked maths at school'...

          Yes, there are a few exceptions. But generally physicists have been good at concepts and working through the mathematics, and mathematicians have been good at inventing new and useful techniques (often centuries before they're needed by physicists).

          As a mathematician, I have to say that although most of the physicists I know are capable of using maths, they rarely understand mathematical beauty and truth, which is essential to be able to come up with new theories and branches of the subject. Of course, for my part, I'm not so good at 'seeing' the physical explanation behind a mathematical equation (e.g. differential equations that describe heat flow through a sheet of metal).

          That all said, S Wolfram appears to be a genius of some sort, so it wouldn't surprise me if he had both mathematical insight *and* physical insight.
      • You have no real clue on Wolfram because you obviously haven't read the book.

        YOU ARE NOT QUALIFIED TO COMMENT if all you've read are the reviews. So please shut up until you read the book.

        For only $45 from Amazon, A New Kind of Science is physically one of the best deals I've ever seen in a book. Its size is enormous (well over 1200 pages), and the quality of the paper, binding and printing process is extremely excellent, because the high resolution illustrations required it.

        "Many of the pictures in this book have a rather different character from things that are normally printed. For unlike traditional diagrams consisting of separate visible elements -- or photographs involving smooth gradiations of color -- they often for example contain hundreds of cells per inch, each in effect independently black of white. And to capure this properly required careful sheet-fed printing on paper smooth enough to avoid significant spreading of ink." ... "The book was printed on 50-pound Finch VHF paper on a sheet-fed press. It was imaged directly to plates at 2400 dpi, with halftones rendered using a 175-line screen with round dots angled at 45 degrees. The binding was Smythe sewn."

        Even if you never read this book and only use it as a paper weight or prop to pick up girls, it's still the highest quality paper weight or chick magnet you'll ever find for the money. If Springer-Verlag had published A New Kind of Science, it would probably cost at least $250, be printed on cheap K-Mart toilet paper, and they wouldn't have even considered putting a fresh ribbon in the typewriter.

        If you do bother reading the book before trying to write a review or refute its contents by personally attacking the author, it will certainly change your view of the universe.

        -Don

        PS: Here's a dynamic cellular automata snowflake generator that I wrote a while ago, inspired by Margolis and Toffoli's "Cellular Automata Machines: A New Environment for Modeling [amazon.com]" [MIT Press, 1987]:
        AethOTron: http://www.DonHopkins.com/AethOTron [donhopkins.com]

      • You accuse Wolfram of 1) having no real clue of math, and 2) obviously not having grasped the usefulness of a clean mathematical formalism.

        So have you ever heard of a widely-used product called MATHEMATICA?

        Open the URL http://www.mathematica.com [mathematica.com], notice where it redirects you to, learn about it, and see how laughably wrong and totally off-base you are in your accusations that Wolfram doesn't understand math.

        Krapangor, I find it impossible to believe that you know much about math yourself, if you've never heard of Mathematica. But for you to say that Wolfram doesn't understand math -- that takes the cake! Ha ha ha!

        -Don

      • Krapangor writes: "However Wolfram doesn't seem to understand the complexity which arises even from continuous systems and that in fact non-continuous dependencies can turn up in continuous systems. Do I even have to mention the Lorentz system at all, everyone should know it. But he is just a physicist after all."

        This is from the index of Wolfram's "A New Kind of Science":

        Lorentz, Hendrik A. (Netherlands, 1853-1928)
        and relativity theory, 1041
        Lorentz contraction, 1041
        Lorentz gas, 1022
        Lorentz transformations, 1041, 1042
        Lorentzian spaces, 1051

        From the notes for Chapter 9, refereing to Page 522, History of Relativity, on page 1041:

        [Mentions Galileo, Newton, Maxwell, Michelson, Morley, FitsGerald...] Already in 1904 Lorentz pointed out that Maxwell's equations are formally invariant under a so-called Lorentz transformation of space and time coordinates (see note below). [Mentions Einstein, Minkowski, Mach...]

        Yet as I discussed earlier in the chapter, if a complete theory of physics is to be as simple as possible, then most things like relativity theory must in effect be derived from more basic features of the theory -- as I start to try to do in the main text of this section.

        [End of quote from Wolfram's "A New Kind of Science".]

        How about reading the book before dismissing it by insulting all physicists?

        -Don

      • I suppose Wolfram, being a physicist and therefore having no real clue of math, means that a model consisting just of smooth operator equations won't be sufficient.

        Actually Steve was not really a physicist per se, he was a mathematician/computer geek type who happened to hang out with physicists and dabble in their field while making use of the ludicrously lavish resources that particle physicists have access to. He was at the Rutherford labs about five years before I worked with the people there in the same sort of semi-detached role.

        In developing Mathematica Steve pretty much worked the field of mathematics. To call him 'only a physicist' sounds to me like someone trying desperately to promote themselves by putting others down.

        Where people can legitimately ask what Wolfram has been playing at is his stweardship of Wolfram Research these past ten years. Back in 1994 a whole new version of Mathematica came out that was very close to being a Web browser. I talked to him about something in that line, he got al excited and... nothing happened. It is clear now that he missed the Internet explosion while he was writing the damned book.

        Where Wolfram Research is really vulnerable is the ridiculous cost of their product. If you thin MSFT price gouges compare the price of Excel and Mathematica. If someone coulf work out a way to graft SMP functionality into a spreadsheet style interface they could take Wolfram Research appart.

    • Saying that Wolfram's book is focused on promoting cellular automata as the way to describe the universe is a gross underestimation of its ideas. From the first few chapters he makes a point to highlight the _equivalence_ of CA and existing mathematical constructs. The real kicker of the book is that using the analysis historically applied to the CA, we can demonstrate previously ignored features of the existing mathematical models.
      • So SW is essentially introducing a new tool for the analysis of existing data. But CA have been around for a long time.

        So his revolutionary advance in the field (which gets him compared to Darwin!) is to suggest that scientists try using a underused tool to fill the gaps in their existing knowledge?

        I must be misunderstanding, because I am a bit underwhelmed.
        • I think your description is exactly what he contributes to the field, when you take away the grandiosity, pomposity, and credit-stealing that Wolfram does. And indeed, it is underwhelming because it's not new scientifically, and it's not new meta-scientific commentary either.
    • I think of computer programming as a means of encoding logical processes, not just a means for computing equations. Therefore, computer programming is much more powerful than mathematics and probably more suited for describing the physical world. Wolfram seems to argue that equations are inadequate for expressing the physical world due to the fact that they often cannot be used to exactly describe the outcome of repeatable events since there is always something unforeseen in the application of equations that affects the outcome. This sounds fair to me, as equations (in physics) are after all, only mankind's attempt to model observable behavior. Given that mankind develops these equations in an evolutionary manner, it's a fair bet to say that that we don't get it completely right. Do we always apply the correct or relevant equations? The question is: Does CA as Wolfram advocates, present a better means of describing the Universe?

  • Has anyone here ever heard... (Score:3, Insightful)

    by Krapangor ( 533950 ) on Sunday June 16, 2002 @10:26AM (#3710913) Homepage
    about a serious scientist claiming that his theories would replace the entire physical model of the universe ?
    Some geniuses did such work, but I never have heard anyone of them making such claims without the in-depth review of others. I must admit I've never heard of any genius exaggerating his own theories so much at all.
    Some people say that's a relatively sure sign for being a crackpot.
    • Isn't it possible that he's such a unique guy that he doesn't fit any kind of mold you're aware of?

      • Re:Has anyone here ever heard... (Score:5, Insightful)

        by gilroy ( 155262 ) on Sunday June 16, 2002 @11:44AM (#3711193) Homepage Journal
        Blockquoth the poster:

        Isn't it possible that he's such a unique guy that he doesn't fit any kind of mold you're aware of?

        Of course it's possible. It's also possible that he's a complete crackpot who, by dint of owning a publishing company, gets to blare his name across the ether.


        Luckily, after millenia of history and centuries of struggle, we've managed to evolve a system that -- much more often than not -- functions to separate the truly original and productive thinker from the truly original and marginal nutcase. It's a system that, amazingly, allows us to make confident statements about things of which we cannot have direct knowledge and that provides relatively surefire ways to establish tests to enhance that confidence.


        That system of course is the system of peer review matched with rigorous experiment, coupled to independent replication of significant results.


        Since the scientific system excludes certain types of claims and certain ways of making claims, it logically runs the risk of excluding the bona fide true revolutionary.... Yet in truth it does not seem to do that all that often. If a result is radical and useful, it eventually works its way into the community. Einstein's theories were nothing short of the demolition of the prevailing, overwhelmingly successful Newtonian worldview. But he made that revolution within the system, and the system accommodated it.


        Too few people appreciate the astounding success and use that follows from a simple, oft-misunderstood fact: Science is not about "discovering truth". It's about quantifying ingorance ... bounding the unknown so as to make it slightly more comprehensible.


        In science we don't know all that much, compared to the vast possibilities of the Universe. But what we know, we know well.

        • we've managed to evolve a system [of peer review] that ... separate[s] the truly original and productive thinker from the truly original and marginal nutcase

          Which is a system functioning in a separate technology realm from industry and invention. I can't directly site the MIT study, but the result is effectively (my analogy) what's seen in child-development. Before the development of a set of social / communication skills small children will play adjacent to each other and rarely interact.

          Neither of these systems (academia / industry) in practice holds the other in particularly high regard. In fact a small fraction (ca 1-5%) of engineers / scientists stay current with what's happening in 'that other area', these individuals, termed 'gatekeepers' are repsonsible for nearly all technology transfer.

          the scientific system excludes certain types of claims ... it logically runs the risk of excluding the bona fide true revolutionary.... Yet in truth it does not seem to do that all that often.

          I guess it depends on what you consider 'often' and 'revolutionary'. Lynn Margulis's discovery that Eukaryotic (all higher order life) cells resulted from the symbiotic relationship between prokariotic cells and viruses was actively derided in biology for a decade.

          Scientists who choose not to live in the arena of academia, or corporate R&D are often the innovators who bring the most real innovations to light.

          Examples:

          • James Lovelock (inventor of gas-chromatograph tools, responsible for Gaia hypothesis and warning of te HCFC / Ozone problem)
          • Itzak Bentov (one of 2 principal inventors of angioplasty and related less-invasive medicine one of the founders of Boston Scientific (now $2B+ sales)
          • Stephen Wolfram
          The common theme among these individuals is that they pursued new work in part outside of established doctrine, and to some extent this was precisely possible becuase they worked outside of 'peer review'

          Lovelock observed in his original book about Gaia that some kinds of research will never be taken on in academia (or the results of completed work will be rejected) because of purely social considerations. He cites the mis-evaluated concerns for safety in nuclear energy, comparing it to the actual (larger) magnitude of toxic chemical contamination risks.

          For a similar example read (or google for) "Brain Sex", a summary of research documenting differences in male and female brain structure. Researchers in this field have uniformly found that because it is not 'PC' to observe that male and female cognigtion / brain structure exhibit meaningful differences, their (almost certainly valid) works are very slow to be funded or accepted.

          These individuals and fields demonstrate how sometimes truly groundbreaking work can only happen outside of the established context. In these instances and many similar ones this happens when an individual can fund his(her) own work and therefor work outside of the peer review system of science.

          Einstein's theories were nothing short of the demolition of... Newtonian worldview

          Actually, Poincare noted the implications of both Relativity and Quantum Mechanics a couple of decades before Einstein applied the mathmatics necessary to fully illuminate the problem.

          'Science' often believes the myth that it is an objective undertaking, not subject to whim or 'current fashion'. Most people who work very long in scientific fields discover that there are (wrong) articles of faith which become codified in 'the literature'. In fact 'Science' is a very human endeavour.

          If peer review and scientific method alone were sufficient to accomplish all new work the examples above would not be true. They may be the exception, however they are clearly (IMHO) important exceptions.

          Whether through introducing new understandings which would have otherwise been missed or effectively bringing new ideas and tools into the marketplace / policy, these are examples of where 'Science' as an institution comes up short.

          None of which, by the way is intended to deny the validity of the various methods. 'Science' progresses through combinations of insight and hard work. Whether the hard work part is practiced to adhere to the rigors of peer review, or to bring an genuinely new idea to market in a form that works, the process is similar.

          • Blockquoth the poster:

            'Science' often believes the myth that it is an objective undertaking, not subject to whim or 'current fashion'.

            That's certainly the middle-school version of it, enshrined in textbooks and handed down as wisdom. As a physics teacher I do my best to work against the myth that science is not about people. But almost every single deconstructionist/revisionist in the field of science sociology makes the equally unwarranted leap to the statement that science therefore is just subjective with no special claim on truth.


            This, of course, is bull-crap.


            Science is a subjective endeavor that leads to objective truth. While there are trends and fashions in science -- because scientists are humans -- the process of peer review and independent replication do move us closer to the truth. Or, at least, they push back the bounds of ignorance, which is much the same thing.


            Even the most outlandish theories can gain acceptance, if the evidence bears them out. It can regrettably take a decade or two, sometimes even longer. But every example you offer indicates the strength of the peer review process, not its weakness.


            What use is it if a lone wolf "gets it right", if we can't tell that he/she got it right? Peer review is an overwhelmingly successful mechanism for weeding out the wrong and discovering the right. Due to the human nature of the participants, sometimes the glorious unbiased evaluation of new work is more honored in the breach. But the system does work, because if a crazy theory happens to be right, the evidence will accumulate -- even through "safe" channels -- and eventually, the peer review system will correct itself.


            Of course, as was once quipped, sometimes you have to wait until all the old scientists are dead. :)


            By the way, Poincare could not have "noted the implications of both Relativity and Quantum Mechanics a couple of decades before Einstein". Quantum Mechanics did not even begin to exist until the discovery of the electron in 1897. Indeed, Planck established the ad hoc basis of the field only in 1900(ref [nobel.se]). Einsten published his first papers on quantum mechanics in 1905. I will grant that Poincare saw a lot of the implications of non-Euclidean spaces, a fundament of Einstein's General Relativity.

    • I think that the big problem is that fundamentally Wolfram's point is rather trite. We can all agree I think that physical processes of growth, differentiation, and plasticity rely upon fairly simple iterated rules. Understanding those rules is the really important/difficult part, especially when the system is utilizing protein as the substrate. Pigment patterning in shells and zebras is well understood not because of a deep understanding of CA, but because of a deep understanding of the biological system. This is where Wolframs approach fails miserably. Can he apply CA to a problem like protein folding? We all know that the rules to produce a thermodynamically stable protein exist, but until we know what the rules are no amount of CA diddling is going to help. After a huge effort to understand protein folding somebody like Wolfram will stand astride that mountain of knowledge and say clearly CA can explain it now that the rules are known, its just that the parameter space is so large CA can not _predict_ what those rules will be.

      No I haven't finished the book yet, and I'm beginning to regret spending $45.
  • Not skim it, I mean read it. I was going to pick it up for my vacation coming up, and really want to know if it's worth the effort, or do you end up with that odd "cold fusion" feeling of being fed a bucket of horse crap?

    • Re:Anyone read it yet? (Score:5, Funny)

      by donnacha ( 161610 ) on Sunday June 16, 2002 @10:42AM (#3710970) Homepage

      or do you end up with that odd "cold fusion" feeling of being fed a bucket of horse crap?
      Or, indeed, a multi-dimensional containment field of horse crap.

    • Re:Anyone read it yet? (Score:5, Insightful)

      by Moriarty ( 8835 ) on Sunday June 16, 2002 @11:36AM (#3711154)
      I read as much of it as I could stand. The thing that irked me even more than when he claims that complexity theory has been languishing, since he stopped publishing, was his excessive use of the word 'so-called'. There's the 'so-called Fibonacci Numbers', the 'so-called Game of Life', and the 'so-called prime numbers'. Are we supposed to think that the prime numbers are not really the prime numbers? Or is Wolfram writing some kind of giant patent application.

      Painful as it was, I read most of the book just to make sure I wasn't missing anything. The truth is that he hasn't had a useful idea in the past 15 years. The rest of it is either just wanking, like his speculations on how the laws of physics could be generated by a CA - pure speculation with no way of using his ideas to solve any real problems. Other times he's just plain wrong, such as his idea that natural selection is not the cause of life complexity.

      His reasoning is pretty flimsy going something like this:
      1. One-dimensional CA are as complex as anything produced by two or more dimensions (he shows a one dimensional cross section of the Game of Life and it looks sort of like his beloved Rule 110 CA which is all he needs for proof. Three or more dimensional CA's are not discussed, since he can't print them in his book)
      2. 1D CA's can only be set up that emulate a small set of patterns. This is refered to as following contraints.
      3. Therefore, everything in nature must be fundamentally simple. There is no way for things to be developed that can follow predefined constraints, and hence natural selection has no ability to optimise organisms, and all life on earth is just stuff that was thrown together any kind of mutants you put together would be just as viable, the brain works the same way, yada yada yada.

      I'm going to be sick. I'm glad I returned it, and please don't get me started on the notes!

      • So, in other words, this is going to turn into yet another line item on a Creationist phamplet? Revolutionary new scientist Wolfram definitively proved with hard mathematics that evolution can't happen: but was censored by the scientific community? Sigh...
    • Yes, I read it. (although long, it really isn't that hard of a book -- Wolfram, like the late Stephen Jay Gould, often uses more words than he needs, and besides that the book is double spaced and full of pictures).

      The real problem is that his key Principle of Computational Equivilence is simply asserted. Wolfram believes that nothing in the universe (including quantum computers!) can really be more powerful than his CA's. Maybe that's true, maybe it isn't, but I'm certainly not convinced.
  • His new book is a fun read (although I am only 100 pages into it).

    One thing that I get from the book is more support for the idea that information processing may be more important to the Universe than physical matter.

    Permit a tangent here: a few years ago (July 1999), I went to a Quantum Approaches to Consciousness meeting at NAU. One current popular theory is that matter in the universe is an uncollapsed wave equation with infinite extent until some form of consciousness observes the matter in question - it is the act of observation that collapses the quantum wave equation.

    Anyway, interesting ideas that are supported by many in the physics community (my Dad is a physicist and member of the National Academy of Sciences, and a few of his aquaintances presented papers at the quantum consciousness meeting) and worthy, I think, to at least not be tossed in /dev/null.

    Back to the topic: I suspect that Wolfram's book will not drastically change the world of science, but it is fun to read.

    -Mark


    • I find it very interesting that a quantum theory invoked what is a common philisophical idea: that at some point the universe ceases to exist as we know it when one reaches a new plane of existance.... the sort of end id say you get when you realize exactly how the universe works and the whole equation collapses.
    • I went to a Quantum Approaches to Consciousness meeting at NAU. One current popular theory is that matter in the universe is an uncollapsed wave equation with infinite extent until some form of consciousness observes the matter in question - it is the act of observation that collapses the quantum wave equation.

      While this is not my field, it is close (I have published in both quantum mechanics and biochemistry and my PhD is in biophysically related field), and I would caution interest in so called quantum consciousness. Not because it is necessarily wrong, but because many of those who believe it want to believe it so much that they are incapable of changing thier mind. (The cold fusion field has similar zealots)

      -Sean
      • I would caution interest in so called quantum consciousness.

        Agreed. I've run into some of those people. They get annoyed when you ask questions like "if the brain uses some big field for internal intercommunication, why don't people get interference when they put their heads close to each other".

        On the other hand, biological brains perform better than we'd expect from the known number of neurons and the gate delays. We're missing something.

    • "One thing that I get from the book is more support for the idea that information processing may be more important to the Universe than physical matter."

      What kind of a statement is that? What is "information"? Where does it reside? Where is "information" "processed"? What the hell does "important" mean to the Universe? What an absurdly useless statement.

  • And is God a software engineer?

    Are you kidding? You mean you hadn't already figured that Heisenberg's Uncertainty Principle and all its quantum-physical consequences are the ultimate kludge to prevent us from looking too deeply into the fact that the entire universe is a simulation, and liable to be switched off at any moment, 'cos the programmer's got bored with it? ;-)

    • God's Notation Sucks (Score:4, Funny)

      by donnacha ( 161610 ) on Sunday June 16, 2002 @10:55AM (#3711000) Homepage

      And is God a software engineer?

      Well, if he is I refuse to work with his code, not until he comes back and notates it properly.

      • Well, if he is I refuse to work with his code, not until he comes back and notates it properly.

        Ouch *brain hurts*
        You're not a programmer, are you? Code gets commented, not notated.
        Sorry if it's a nitpick, but it really really clashed.

        -

        • You're not a programmer, are you? Code gets commented, not notated.
          Internationally either word can be used, with "notation" considered more professional.

          Apologies for not thinking American, I usually catch those mistakes.

          • Apologies for not thinking American, I usually catch those mistakes.

            Apologies for being American, I was born into that handicap, chuckle. It's the first I've come across the usage of "notation" in place of "comment".

            Internationally either word can be used, with "notation" considered more professional.

            It may be normal in some country, but it isn't common enough to even show up on google radar. I just tried a search on "code" and "notation" and looked over the first 200 results. I couldn't find it used in that manner.

            -

  • Egomaniac (Score:1, Flamebait)

    by skroz ( 7870 )
    I've only made my way througha small portion of ANKOS, and my only conclusion so far has little to do with his theories and ideas. I've concluded that Wolfram is, in fact, a desperate egomaniac, and that 20 years of thinking of no ideas other than his own has left a permanent and irreversable mark upon his mental state. The ideas may be brililant, but the guy SERIOUSLY needs to get some sunlight.

  • And is God a software engineer?

    Funny, I thought he was a clock-maker. I guess with this new metaphor, we can be expecting some bug-patches anyday now.

  • The Huessendorff Effect (Score:5, Insightful)

    by MattJ ( 14813 ) on Sunday June 16, 2002 @11:26AM (#3711103) Homepage
    Summary: In computer graphics, there are times when you prefer to use bitmapped graphics (PhotoShop), and times when you prefer to use vector graphics (Illustrator). The first is like a CA, the second is like an equation. Wolfram is so excited to have learned about MacPaint, that he insists we use only it, and drop MacDraw from our toolbox.
    ---

    Wolfram's book is an example of what I call the Huessendorff Effect, which states that in order for an author to create a significant buzz, your book has to make an overly-broad, sweeping claim. E.g., if you want to get the cover of the NYT Book Review, don't take an issue and examine all the evidence and present a nuanced assessment of it all. Instead, make a grandiose statement to get people's attention. After you've got them, they'll read your book, subconsciously discount your hype, and still be left with the impression that you said something "great".

    (Why is it called the Huessendorff Effect? Because I'll never get this phenomenon noticed unless I give it a bold name and declare it to be a widespread law, silly.)

    Anyway, that seems to be the view of most people who produce books like this. Although in Wolfram's case, it seems he truly believes his grandiose claims.

    A quick point about his book (which I haven't read yet, but I've followed his work in cellular automata since 1986). I think he mistakes his mental model for the world with the way the world really is. He applies his CA interpretation to all situations, even when other, simpler models exist. Take hydrodynamics. Everyone looks and sees the same eddies and currents. Now, one person might think of the traditional, nonlinear equations describing that, and another person might think of a cellular automata grid generating that behavior. They are just two ways of describing what's going on, and neither is the "right" or "real" description.

    Wolfram seems to be so excited that CAs can provide simple explanations of many situations that are extremely hard (or impossible?) to describe in old-school equations, that he insists that CAs provide the "real" description of the world, and should replace equational models.

    But Steven, there is no reason to dump equations that are simple and that work. Kepler's laws work, and they are simple. Show me a CA that can describe a planet's orbit in one sentence of rules, because I don't believe it. And even then, show me that your CA grid isn't introducing tremendous round-off error.

    Take-Away Point: In computer graphics, there are times when you prefer to use bitmapped graphics (PhotoShop), and times when you prefer to use vector graphics (Illustrator). The first is like a CA, the second is like an equation. Wolfram is so excited to have learned about MacPaint, that he insists we use only it, and drop MacDraw from our toolbox.
    • How is his CA theory novel exactly? If I automate the solving of a series of Markov chains or PDEs for genetics or insurance purposes, how are the fundamental equations less valid?

      Is he saying that because phenomena are more simply modelled using a CA than a system of equations, that the CA approach is a "truer" representation of the phenomena?

      The articles give the impression that he thinks all equations can be found to be more simply replaced by CAs. Is that really in the book? I will gladly accept your Takeaway Point, but does SW really suggest that.
    • A quick point about his book (which I haven't read yet, but I've followed his work in cellular automata since 1986).

      I was waiting for this part. Translation: I can't be bothered to get off my can and read the book but here's what I think he says so I'll throw out my own half assed opinion.

      Wolfram seems to be so excited that CAs can provide simple explanations of many situations that are extremely hard (or impossible?) to describe in old-school equations, that he insists that CAs provide the "real" description of the world, and should replace equational models.

      Hmmm. Had you actually read the book you would have noticed that he explicitly says that CA are another way (and certainly not the only way) of modelling problems in the same way that Partial Differential equations and Lagrangians do. What he does say is that they produce better and simpler models that describe the phenomenon. The map is not the terrain but a map is useful.

      Thanks for the brain-dead summary though -- I don't know where the hell you came up with that analogy...

    • . Take hydrodynamics. Everyone looks and sees the same eddies and currents

      Show the world an elegant solution for a viscous, unsteady 3 dimensional fluid dynamics problem and the world will award you a nobel prize. Actually you will probably get several, because your work will revolutionize every industry in the world (not to mention make every math textbook ever written instantly obsolete). At the very least you'll make the cover of AIAA journal.

      i don't know if you can do it using "cellular automata" (doubtful), but you sure as heck can't do it with calculus, and even with half the variables assumed away, finite element analysis/computational fluid dynamics makes supercomputers cringe. If Wolfram's presentation of these ideas brings us a tiny bit closer to solving these kinds of problems, he deserves every accolade.
  • Much of what has been written about Wolfram's new book has focused on his lack of acknowledgements and use of the first person. While I don't necessarily agree with his approach, like many others, I do agree with many of the arguments that he (and others before him) are making. On the other hand, perhaps his "ego-centric" approach to promoting these theories is exactly what's been needed, and may even get those truly deserving of recognition more of it in the end.

    So far, complexity science etc. is mosty ill-defined and includes a number of difficult concepts - not only for the general public but also for those people in the very fields it could make the most impact in. The ego-centric hype and style generated by Wolfram gives the reader a sort of personification of the theory to use as a reference point, with Wolfram acting as a protagonist to help the reader through an otherwise "complex" plot - not disimilar to the use of Brian Arthur in "Complexity" by Waldrop (also controversial in some circles).

    Also, articles about Wolfram are a hell of a lot more interesting to the media and the public (and even many academics) than articles just about complexity. This includes the whole Wolfram myth and the controversy over "his" theories, but also includes the controversy over his ego often through quotes from individuals like Seth Lloyd and Ray Kurzweil or a discussion of the accomplishments of the Santa Fe Institute etc. Can you imagine the book or the theory (or the other people being cited) getting as much coverage in the NY Times etc. if it were a collection of peer-reviewed physics papers or a well-referenced text-book style overview of the state of the art in cellular automata?

    Using this approach (along with pretty pictures and aggresive pricing) Wolfram has probably generated more interest in complexity science, cellular automata, digital physics, and perhaps even scientific methodology and notation among more people than any previous proponent of these ideas. Now, popularity does not good science make, but what he is essentially proposing is a wholesale shift in scientific notation - a bold idea that will need broad consideration to work - but also an idea that is inherently unsexy to most and probably needs a bit of showmanship and publicity (good and bad) in order to get the discussion really cooking.
  • Who says god couldn't have implemented his CA in hardware? That would make him some other kind of engineer (I'd say electrical, but there's no telling what god might have implemented the world in. The way this place runs, you'd think it was a difference engine).
    Then there's the old saw about how god must be a civil engineer. As they say (reffering to the human reproductive system) would have put a waste disposal system in the middle of a leisure area?

  • There are 80+ reviews on amazon.com, some of which are detailed and specific. The overall rating is running at 3 stars out of 5, with most people apparently of the opinion that writing a long scientific work after living in hiding for ten years is a very bad idea. The gist of the complaints is that he's rather full of himself, tends to take credit for the work of others, and enjoys the "non sequitur" approach to scientific discovery.

    The people giving it 5 stars or 1 star seem to be doing so after simply reading the back cover.

    I ordered a copy anyway. If it causes this much controversy it's bound to be interesting. :-)

    • More accurately, it's a bad idea to go into hiding for 13 years to write a book that's based on the work of a lot of other people, using the Internet as your primary research tool and then only marginally acknowledging sources in ridiculously long notes at the end.

  • Wired has a few articles. (Score:3, Informative)

    by dodald ( 195775 ) on Sunday June 16, 2002 @11:41AM (#3711181) Homepage
    This is a fairly long, but a good read [wired.com]

    There was one with pictures, but I can't seem to get the page to load (Search "Wired News" for wolfram, the article called "A New Kind of Thinking")

  • Go to the Amazon page [amazon.com] and click "See all customer reviews" [amazon.com]. The earliest reviews--the ones written by reviewers who got prepublication copies--are all very positive. The later ones largely brutalize the book. Reading through them (along with the mixture of positive reviews written by gullible people) gives you a pretty clear picture of what's happening.

    I've looked at the book for a while. Yes it has some interesting stuff in it. It doesn't come anywhere near the claims the author makes for it. My favorite review on Amazon is titled "A new kind of review", posted May 20. You'll have to scroll quite a ways down to read it, but it's very funny. After you read that review, you'll understand completely.

  • A book crying out for an editor (Score:5, Insightful)

    by mrneutron ( 61365 ) on Sunday June 16, 2002 @11:49AM (#3711210)
    I bought the book, and am reading it now. It's painful to read, despite the fascinating ideas contained within.

    He states 'complex behavior arises from simple behavior', or words to that effect, thousands of times. It's mind-numbing.

    It's clear he surrounded himself with sycophants who couldn't (or wouldn't) tell him to trim this beast down. The book could easily be 1/4th its current size and deliver the same information.

    He also takes credit for discovering cellular automata repeatedly, despite the fact that this concept has been known for at least 50 years. The Game of Life was first programmed over 30 years ago.

    There's a section where he computes primes using cellular automata ala the Sieve of Erastosthenese (very cool), and I half expected him to take credit for that idea as well.

    The question is: do his findings outweigh his egomania? For me, I'm still reading.

  • Submitted to Amazon on Friday:

    ***** = A massive acheivement
    Reviewer: James Mitchell from San Francisco, CA United States

    Stephen Wolfram's 'A new kind of science' is a massive achievement, true to the grandest traditions of self-publishing. In this stunningly beautifully laid out tome, Wolfram displays the fruits of long, intense dedication to the most sublime hubris, progressing from what are genuinely intriguing results of arduous empirical research to sweeping delusions of competance. Who would have thought that you could replicate the behavior of fluid movement on a computer by explicitly modelling it? Stunning, indeed. Doubly stunning is that a work of such gargantuan inanity could contain such a concise and lucid explanation of chaotic processes. That a mind that can so clearly explain the phenomenon of super-critical dependence on initial conditions can also produce so much excess self-congratulation for producing such a vapid vehicle for presenting this work will provide excellent working material for budding young doctoral candidates in psychology.

    In its best movement, Wolfram's book spends a great deal of time demonstrating how his computational artifacts are unable to work out the results of constraints, in the process demonstrating the total futility of over a decade's worth of research. Amazingly, Wolfram presents the inapplicability of his work as a mark of its virtue -- that you can produce totally unpredictable and incomprehensible behavior without regard to the actual process one is researching. Given an intellect of such Colossal stature as Wolfram's, this massive tome is in and of itself the most solid, bulletproof example of the value of peer-review. That such a Herculean effort by such a gifted mind could produce a work of such stunning irrelevence should dissuade even the most ardent researcher from removing himself or herself from the academic community.

    In addition to being the absolute paragon of case-studies for the value of peer-review, the book is also physically beautiful; it's rich yellow and black artwork will spice up even the most pedestrian of bookcases.

  • The metaphor for God (Score:5, Insightful)

    by gilroy ( 155262 ) on Sunday June 16, 2002 @11:52AM (#3711224) Homepage Journal
    Is God a software engineer? Not any more -- or any less -- than He is a watchmaker.


    Here's the overriding truth of worldviews and metaphors: For at least the past five hudnred years, we in the West have taken the dominant mode of industry and "explained" both human consciousness and the Godhood in terms of it.


    First, of course, industry was agriculture... and God was basically a farmer, creating and tilling the Earth, making it ripe.


    Then we came upon clockworks. (Too many miss the deep pyschological impact that the idea of time-keeping had upon the world.) Nice orderly systems that run more or less regularly, mimicking the order seen in, say, the motion of planets. And here, of course, God is the ultimate watchmaker.


    The Age of Steam comes next and now God is the ultimate civil enginner. The Universe is a vast and complicated -- but ultimately comprehensible -- machine. It's made of discrete little bits that fall into recognizable types. If we understand the types and how they interact, we can reverse-engineer the machine.


    Now we're in the Age of Information. The rising dominant archetype is the digital computer, revolutionizing our world the way that the steam engine did the 1700s. It almost goes without saying that of course some people are going to see digital computers in everything -- even the deepest bits of the Universe -- and so of course someone is going to claim God is the ultimate software engineer.


    My impression is that these metaphors reveal less about God than they do about us... we don't come any closer to understanding God through them, but we might -- if we pay attention -- come closer to understanding how we understand ourselves.

    • I agree very much. Newton's dynamics and gravity, and Maxwell's electromagnetism are simply models of the world. They are 'phenomenological' theories that can predict the outcomes of many physical situations, but are totally agnostic as to what is happening within.

      Einstein's theories of relativity that combine the above, are more accurate and elegant (i.e. conceptually simpler) than the two. But the more accurate predictions do not mean that the model is any closer to the 'real' workings of nature.

      Wolfram's model may be even more accurate, but there can never be a conclusive proof if it really reflects the reality.

      I remember a lecture by Benoit Mandelbrot I attended a few years ago. He showed the exactly same idea as Wolfram is explaining, that starting from very simple algorithms you could iterate many natural patterns. What really struck me was Mandelbrot's note on the idea of patterns themselves: "Are there patterns out there in nature, or are the patterns only in our heads?"

      The latter possibility comes back to what you've explained, that the model tells more about the current society, than it does about nature. Of course, the question looks like it can never really be answered.

      • This point is certainly underappreciated. What people like Neweton and Einstein (idealy) do is develop _models_ that help us understand the conventions of the natural world. They allow us to explain and predict its behavior in terms of (hopefully) successively more accurate terms, figuring out which elements are important and which are not.

        But when it comes to the natural world itself, we can't with any deal of assurity prove even our most basic assumptions. One of the most startling of these is causality: the idea of causality is quite entirely a conceptual idea: there is no way to actually prove that one event "causes" another: only that they are correlatively linked by a particular relation or supposed mechanism.
    • God?
      We made him.
      In our image.

      We need something to believe in. Whatever people can believe is ok.
      I for one would prefer that more people believed in science. But science takes a lot of work, and physics doesn't claim to hold all the answers, yet, or ever.

      So most people still like smoke and mirrors better.
      And, really, who can blame them.
      Easy answers, set rules, authority from above.
      It is very seductive.
      • Blockquoth the poster:

        So most people still like smoke and mirrors better.
        And, really, who can blame them.
        Easy answers, set rules, authority from above.
        It is very seductive.

        I think you have to be careful about too easily dismissing the religious impulse. I myself am not religious in any traditional sense of the word, but I know far too many clear-headed, rational people who nonetheless believe in an ineffable Other. Not everyone runs to religion to run away from responsibility.


        Just like, not everyone runs to science to achieve rationality.

  • A page of links to ANKOS reviews... (Score:3, Informative)

    by jnana ( 519059 ) on Sunday June 16, 2002 @01:37PM (#3711580) Journal
    See here [usf.edu] for a page that links to about 15 reviews of ANKOS. My favorite is this review [maa.org] for the Mathematical Association of America [maa.org].
  • And all along we thought God screwed up by making the world with all these problems.

    Maybee he just wants to see then end result too.

  • Two Kinds of Science (Score:3, Interesting)

    by Royster ( 16042 ) on Sunday June 16, 2002 @04:58PM (#3712057) Homepage
    There are, generally speaking, two kinds of science.

    Some sciences are descriptive and others are predictive. Physics is the ultimate in predictive science where temendously precise pridictions
    about interactions can be calculated in advance. But there's a limit to what we can reasonably calculate. Many problems, like a Newtonian
    3-body problem, have no closed solution and require numeric approximations to calculate anything. Other problems exhibit sensitivity to initial conditions and result in chaotic behavior. Precise predictions are no longer possible.

    Other sciences are descriptive. They attempt to classify and organize observations into meaningful systems. Cladistics, pre-Darwin, described anatomical similarities between known species. Eventually, the resulting family arrangements were understood as evidence of underlying evolutionary processes whereby closely related species were
    only recetly divergent and species with greater differences were less closely related.

    Wolfram offers us a little of both in A New Kind of Science (which I have bought, browsed, but not yet read in depth). Only time will tell
    if the systems he's calaloged will pay off in other disciplines. It may very well be that, once we know what to look for, natural analogues
    of his systems may be all around us waiting to be discovered.

Slashdot Top Deals

It isn't easy being the parent of a six-year-old. However, it's a pretty small price to pay for having somebody around the house who understands computers.

Close