Kuro5hin.org: technology and culture, from the trenches
create account | help/FAQ | contact | links | search | IRC | site news
[ Everything | Diaries | Technology | Science | Culture | Politics | Media | News | Internet | Op-Ed | Fiction | Meta | MLP ]
We need your support: buy an ad | premium membership

[P]
Review of Stephen Wolfram's A New Kind of Science

By AzTex in Op-Ed
Thu Jul 18, 2002 at 03:12:14 PM EST
Tags: Books (all tags)
Books

I looked forward to reading Stephen Wolfram's "A New Kind of Science" for a number of reasons. First, his company's program, Mathematica, was one of my favorites back in grad school. Second, I had not read anything that stimulated my mind in some time. Finally, somehow the book had risen to be a top seller on Amazon and is apparently still back-ordered. Could it really be that good of a read?


Reading the Preface and Chapter One, you may notice that Wolfram makes extensive use of the word "I" and of the phrase "this book". In these sections, he also takes the opportunity time and time again to tell us just how long it has taken him to write "this book" and how much effort it has taken him to write "this book". Should we be impressed with the author for his labor?

The author also promises that his discoveries will revolutionize understanding in Mathematics, Physics, Biology, Social Sciences, Computer Science, Philosophy, Art, and Technology. Wolfram is very ambitious.

In Chapters 2-6, Wolfram shows many examples of his experiments with cellular automata, mobile automata and other repeating algorithms and the sophisticated patterns that these algorithms can generate. These patterns are interesting and impressive and Wolfram presents this material in a very clear and concise manner.

Although it is a great deal of material, a reader should go through all of it to really appreciate the potential of a simple repeating algorithm. In these chapters, Wolfram shows himself to be an excellent writer. The material is understandable with only a basic understanding of mathematics or computing.

However, work in this field has been done before. Cellular automatons have certainly been studied in biology. He makes passing mention of other papers but provides no explicit references. In the notes for Chapter 2, he does list a number of his own papers related to the subject. But he does not state when and where they were published, if at all! If he spent so many years, working on "this book", why did he not at some point take the time to provide a clear reference to the works of himself and of others in this field?

Frankly, I was left wondering what is Wolfram's original work and what has been done before. I was also dismayed that there is no peer review in evidence. I certainly hope that his "New Kind of Science" will include references and peer review.

To be fair, Wolfram makes some mention other's work, albeit without references, in the notes for Chapter 2. The author labels these as "close approaches" to the discoveries that he himself has made. The motivation for listing these close approaches may be for the opportunity to point out that how and why others failed while Wolfram succeeded.

From Chapter 7 to the end of the book, Wolfram uses the outcomes of the experiments of earlier chapters an an attempt to relate simple algorithms to natural phenomena. But most of it is hand-waving.

He appears to make some headway with crystals and simple examples of plant growth. He shows photographs of snowflakes, leaves and stems and compares them somewhat favorably with the patterns produced by particular cellular automations. There are, of course, no exact matches.

The cellular automation results certainly remind the reader of snowflakes and plant stems but they are not mistakable for anything but cellular automation results.

If this new kind of science has any promise, then the author has missed an opportunity to demonstrate it by comparing say, a series of snowflakes to a cellular automations which are indistinguishable from the respective snowflake. Such an exercise would have required a great deal of computing power but it would have been worth it for the resulting triumph.

Continuing with biological examples, Wolfram shows a number of photographs of common seashell and animal shapes. Curiously, he does not attempt to compare these to any cellular automaton or other results.

Over the course of discussion of real biological structures and pigmentation patterns, Wolfram makes it clear he does thinks that natural selection does not explain the observed patterns of biological systems. But he does not actually state why he holds this opinion. He says that he thinks that cellular automata algorithms are being chosen at random and natural selection has nothing to do with it.

From the tone of Wolfram's discussion of the role of natural selection in biology, he obviously is not well versed in either.

I suggest that a more reasonable suggestion would be that when a single organism is growing, cellular automata algorithms determined by alleles are used to perform tasks like the actual construction of shells or distribution of pigmentation. But this does not mean that natural selection is not at work driving the overall adaptive fitness of the species.

Use of certain cellular automata algorithms may offer a greater chance for individual survival over other algorithms and mutation may alter genes thus introducing new or modifing existing algorithms.

Wolfram moves on to a discussion of the role of cellular automata in financial systems. But I will not comment on his financial discussion because, unlike Wolfram, I do not talk about things that I do not understand.

In Chapter 9, Wolfram takes on fundamental physics. He starts by using cellular automata to model a 2-dimensional ideal gas.

The actual point of these models is to demonstrate the behaviour of cellular automata in the context of entropy and The Second Law of Thermodynamics. It is not clear to me what Wolfram is getting at. None of these models involve heat, temperature or work. He keeps mentioning randomness and changes in complexity instead of simply performing a calculation of the changes in entropy of these systems.

On page 453, Wolfram claims that one of his cellular automation models violates The Second Law of Thermodynamics. Of course, without a simple calculation of change in entropy, it is not clear why it is in violation. He provides no calculation showing a decrease in entropy over time. Then based on the supposed violation by this model and some other irrelevant discussion of order and randomness, he states on page 457 his Earth-shattering conclusion that The Second Law of Thermodynamics is not valid.

There are no observed instances in nature of a violation of The Second Law of Thermodynamics. Most reasonable people would say that behavior exhibited by a model that is contrary to what is observed in nature indicates a flaw in the model. But Wolfram is saying that the supposed violation exhibited by his model is more important than observation.

Stephen Wolfram, if you are reading this, I have two words for you:

  • Crack.
  • Pot.

More seriously, I hate to quote trhurler but in this case it may be relevant, " The greatest conceit, and the greatest mistake, is to believe that your model IS what is modeled."

At this point, I did not judge the remainder of the book to be worth reading. I skimmed the rest of it and picked up enough about his Principal of Computational Equivalence to see that he is again rehashing other's ideas. The Universe is a computer.

In summary, if you have this book on order, see if you can cancel your order. If you can return your copy unread, do so.

A New Kind of Science was published in 2002 by Wolfram Media, Inc.
Hardcover: 1192 pages.

Sponsors

Voxel dot net
o Managed Hosting
o VoxCAST Content Delivery
o Raw Infrastructure

Login

Poll
The title of "A New Kind of Science" should be changed to:
o Science is Dead; Long Live Science! 15%
o A Work of Genius by a Genius 7%
o "This Book" 17%
o The Holy Bible, Stephen Wolfram Version 11%
o A New Kind of Nonsense 19%
o The Same Old Nonsense 10%
o Coffee Table Weight, Vol I 17%

Votes: 76
Results | Other Polls

Related Links
o A New Kind of Science
o Mathematic a
o Amazon
o Preface
o tell us just how long it has taken him to write "this book" and how much effort it has taken him to write "this book".
o work in this field has been done before
o The Second Law of Thermodynamics
o The greatest conceit, and the greatest mistake, is to believe that your model IS what is modeled.
o Also by AzTex


Display: Sort:
Review of Stephen Wolfram's A New Kind of Science | 151 comments (132 topical, 19 editorial, 0 hidden)
+1 fp many valid points (4.33 / 3) (#8)
by nodsmasher on Wed Jul 17, 2002 at 11:11:39 PM EST

also quite funny
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Most people don't realise just how funny cannibalism can actually be.
-Tatarigami
worth reviewing, but not this (4.63 / 11) (#9)
by speek on Wed Jul 17, 2002 at 11:23:41 PM EST

Wolfram's book is very much worth reading, I'd like to point out. I'm most of the way through it myself, so I'll give some of my thoughts regarding the book and this review:

Wolfram is clearly full of himself. The sheer number of times he says things like "..discoveries made in this book..." (that's a smart book!), "...as I have discovered, great complexity can come from simple rules..." is absolutely staggering. It's something to get past though, not dwell on.

He goes through a lot of examples of simple rule systems. For someone like myself, this was very helpful - and the book is very clearly written in this section. Others may prefer to skim this part, looking at the pictures, reminding themselves of the various types of simple rule systems (cellular automata, mobile automata, subsitution systems, network systems, number systems, etc).

The discussion of biology is the weakest part. However, this reviewer's dismissal of what Wolfram says about natural selection shows only the reviewer's ignorance of the matter. The biggest problem with Wolfram's characterization of natural selection and evolution is that it is not substantially different from the mainstream scientific view. Both, however, are quite different from the popular view of evolution and natural selection, which has misguided notions of "directed", "for a purpose", "greater and greater complexity", and various other teleological confusions.

No, the real problem with the biology section is more complex. Let's say Wolfram's absolutely right and that the universe is, at the very bottom, governed by a simple rule - some sort of network rule, let's say. What does that tell us about systems at the level of biological entities? There are far too many network nodes involved to ever model with a computer, and there's no guarantee that new, also simple, rules would emerge to govern interactions between biological entities.

Take protein folding as an example. Depending on many factors, proteins "fold" in particular ways. This folding is very interesting, because it often gives insight into how the protein will bind to other proteins. If one could predict how a given protein would fold accurately, that would be a marvelous thing. But, even if a single simple rule were at the bottom of the universe, how does that help with this problem? In a protein, we're already too large for any computer to model all the smallest discrete "particles" involved, and I see no guarantee that there would be a "protein folding" simple rule that would emerge. It could very well be the most god-awful complex rule you could imagine, and have to take into account every possible environment that protein found itself in.

Onto physics. Wolfram is much more impressive here (sadly, the reviewer got stuck on his own misunderstanding of the second law of thermodynamics and went no further). If one posits that the universe is a system of networked nodes, updated via a simple rule in time, one finds that certain things like the speed of light as an absolute, and special relativity, become fairly obvious consequences. Bell's theorem is given a simple and lucid explanation. And the apparent randomness of some quantum phenomenon also becomes an expected result.

I had considered writing a review of this book, when I finished, but my main point was going to be that, far from creating a "New Kind of Science", Worlfram may simply be drawing attention to a "New Kind of Tool" for science. And not entirely new, either. The basic idea is two-fold: Since very simple rules can give very complex results (it's important to really look at the examples and ask yourself - would you ever have believed such results were possible from rules that simple?), we ought to be looking specifically for extremely simple underlying rules. The basic reason to favor simple rules vs complex rules in your search has to do with ease of finding them (it's still hard, unfortunately), ease of understanding them, and a basic preference for simplicity and elegance.

The other part of the "tool" is computers, to use as experimental devices. This, they already are, though in a subtly different way than Wolfram would like. He'd like to see the goal be to develop a computer program that describes and predicts, whereas most scientists would like to end up with a mathematical equation that describes and predicts. Wolfram suggests many good reasons for believing that a computer program as model holds more value than a mathematical (particularly calculus-based) equation as model.

There you go, in a nutshell. I'd say it's worth reading - it goes pretty fast.

--
what would be cool, is if there was like a bat signal for tombuck -

Re: worth reviewing,... (5.00 / 4) (#13)
by AzTex on Thu Jul 18, 2002 at 12:11:28 AM EST

speek, Thanks for your comments. But...

You said,

the reviewer got stuck on his own misunderstanding of the second law of thermodynamics and went no further
I really would appreciate it if you could tell me what it is about The Second Law of Thermodynamics that I misunderstand.

Also, to clarify, nothing about The Second Law of Thermodynamics caused me to deem Wolfram a crackpot. It was the fact that Wolfram used the fact that his models make predictions contrary to those of The Second Law to say that The Second Law was invalid. Only if observation is contrary to the predictions of The Second Law should we say that The Second Law is invalid. He thinks that his model is more important than the real world. That's what I call crackpot.



solipsism: I'm always here. But you sometimes go away.
** AzTex **

[ Parent ]
2nd law of thermodynamics (5.00 / 1) (#38)
by speek on Thu Jul 18, 2002 at 08:59:56 AM EST

The second law is an observation made by scientists. There is no deeper understanding of the situation than that's what we see. People like Stephen Hawking talk about the 2nd law, not as though it's a law, but rather as though it's a bit of an anomaly. The 2nd law breaks with virtually all other laws of physics in that it is not reversible. Why does this one law seem to go only one direction in time?

Theories abound, of course - but many suggest that the 2nd law is not absolute. Many, not just Wolfram, suggest there would be times when entropy would decrease. If a scientist develops a model that predicts results contrary to a current understanding, then that is something to be excited about. Hopefully, it is testable and doing those test could potentially create new understanding.

--
what would be cool, is if there was like a bat signal for tombuck - [ Parent ]

2nd law broken? (5.00 / 3) (#45)
by rleyton on Thu Jul 18, 2002 at 10:29:03 AM EST

Ironic perhaps, but I've just seen this article over at the BBC today - scientists in Australia have demonstrated a small decrease in entropy...

Ho hum... I'm waiting for the cold fusion comparisons now... or comments about "everything being upside down in Australia"

--
Ooooooooooooooh! What does this button do!? - DeeDee, Dexters Lab.
My Website
[ Parent ]

interesting (none / 0) (#50)
by speek on Thu Jul 18, 2002 at 11:03:24 AM EST

Of course, every 2nd law advocate can always fall back on the old "well, the law only applies to closed systems - that system was open". All well and good, but nobody knows for sure if the universe is a closed system!

--
what would be cool, is if there was like a bat signal for tombuck - [ Parent ]

Hmmm (5.00 / 1) (#52)
by jmzero on Thu Jul 18, 2002 at 11:09:47 AM EST

A few years ago, a tentative theoretical solution to this paradox was proposed - the so-called Fluctuation Theorem - stating that the chances of the Second Law being violated increases as the system in question gets smaller.

I think this is central to the question.  I've always understood the second law to be a statistical law.

In this particular experiment, it seems to me like it's going to be hard to add everything up.  Their effect is so short.. and how do you measure the entropy of their whole system so quickly.

.
"Let's not stir that bag of worms." - my lovely wife
[ Parent ]

2nd law of thermodynamics (5.00 / 1) (#63)
by bgalehouse on Thu Jul 18, 2002 at 02:07:46 PM EST

The second law of thermodynamics is a derived law, derived from quantum mechanics. It therefore breaks none of the laws of quantum mechanics. One might argue that it breaks the laws of general relativity, but only in the lack of UFT sense that all of quantum mechanics breaks the laws of relativity.

Quantum mechanics is reversable. Well, kinda. It is hard to say if measurements are reversible, because it is hard to say what constitutes a measurement device precisely, and so it is hard to say what measurement devices can be constructed. The ussual Copenhagen interpretation more or less assumes that they aren't made of wave functions. Which makes it hard to decide what they are made of.

But I digress. Wave mechanics is undenyably reversible. But consider friction. At the quantum level, there is no friction. And so friction is just a model - little more than a figment of our imaginations. It will still kill you if it melts your climbing rope, and nobody has figured out how to make it go away in a macroscopic system. Err, unless you count the limited case of superfluids. And similarly I suppose that a superconducting storage ring loaded with energy in the chill of deep space would hawe constant entropy.

But none of these break the laws of thermodynamics as dervied from QM, as by derivation they only apply to systems of typical quatum particles. If Wolfram was suggesting that his entropy free cellular automata would be good models for some sort of strange quantum system, then he might have something to work on in a lab that might theoretically work. But the US patent office requires a working model for patents on perpetual motion, so obviously he shouldn't be yelling too loudly just yet.

[ Parent ]

sorry, you go against everything I've learned (none / 0) (#64)
by speek on Thu Jul 18, 2002 at 02:29:41 PM EST

The second law of thermodynamics has been around lots longer than Quantum Mechanics. I've never heard of anyone "deriving" it. Your example of friction actually illustrates this - showing that the second law is simply a description of what we observe, with little understanding on our part of why or how.

I don't know if you're confused, intentionally spreading disinformation, or so far over my head that I have no clue, but my guess is the first.

--
what would be cool, is if there was like a bat signal for tombuck - [ Parent ]

Boltzmann (none / 0) (#69)
by dennis on Thu Jul 18, 2002 at 04:07:45 PM EST

the second law is simply a description of what we observe, with little understanding on our part of why or how.

Come on. Boltzmann derived it from Newton's laws in the 1890's.

[ Parent ]

good to know (4.00 / 1) (#71)
by speek on Thu Jul 18, 2002 at 04:36:58 PM EST

Ok, thanks for the info. That's a statistical derivation, right? Which does not a law make, and does not rule out occasions of decreasing entropy. Also has nothing to do with QM, far as I can tell.

--
what would be cool, is if there was like a bat signal for tombuck - [ Parent ]

Large numbers (none / 0) (#74)
by dennis on Thu Jul 18, 2002 at 05:08:52 PM EST

Yeah, it is statistical. Law of large numbers, nothing more. Thus the BBC article people are linking to - get the numbers small enough, and the 2nd law doesn't have to hold. For anything you can observe with the naked eye, the numbers are so huge that you'll never see entropy decrease.

Boltzmann predated quantum mechanics; we would still have entropy if we lived in a classical universe. But the second law can be derived from quantum mechanics as well.

[ Parent ]

QM and stat mech (none / 0) (#76)
by bgalehouse on Thu Jul 18, 2002 at 05:34:06 PM EST

While a certain amount of thermodynamics can be derived using classical mechanics, there are certain aspect which fail spectacularly without quantization. I once had a physics prof who though that "Rayleigh-Jeans and the ultraviolet catastrophe" would make a good name for a band.

[ Parent ]
You haven't read any statistical mechanics (5.00 / 1) (#75)
by bgalehouse on Thu Jul 18, 2002 at 05:28:46 PM EST

There is an entire subfield of physics related to the derivation of the second law and refinements to the law. It is called statistical mechanics. I requires knowledge of QM and of statistics, and so it is rarely taught to non-physics students, and rarely described in lay science discussions. But there are lots of textbooks on the subject, etc.

Nothing I've said is particularly contraversial (except maybe the bit about measurements) to physicists. Don't get me wrong, the law did exist before quantum mechanics, both as a empirical law and later derived from classical mechanics in certain cases.

However, today QM is considered to be the reason for the law, and the mathematical derivation of the law from QM is extreamly well understood. Similarly for friction, even if the number of particles involved makes some of the caculations which one would really like to do intractable in practice.

[ Parent ]

ok (none / 0) (#77)
by speek on Thu Jul 18, 2002 at 05:54:53 PM EST

It still seems strange to me that a non-reversible law could be derived from reversible equation.

And relating back to Wolfram, his examples of CA that "violate" the 2nd law don't really violate in a larger sense, but only on small scales.

--
what would be cool, is if there was like a bat signal for tombuck - [ Parent ]

Reversibilit and entropy (none / 0) (#79)
by bgalehouse on Thu Jul 18, 2002 at 06:57:16 PM EST

Consider the RC4 algorithm. This cipher algorithm maintains a state array. The cipher output algorithm is reversible, so no information is lost on each iteration.

However, if we start the cipher output with the state array in a fairly organized arrangement, (suppose that the initialization algorithm works poorly) then over time it will seem more and more random. If we reversed the algorithm, we'd come back to the more organized state, but the odds of happening upon an organized state normally are quite low.

Now, the fundamental reason for the second law is that 'having energy available' generaly translates into being 'non-random', or 'much less random'.

So, if you had a boiling pot, turned off the heat, let it cool down, and then somehow reversed time, the liquid would heat up again, drawing heat out of the air, etc. Everything is lined up perfectly for this, because it was boiling before you turned off the heat. Most arangements of water and air will not spontaniously heat up though. In fact, the odds of finding such an arrangement by chance are so low that it is unlikely to happen even once in the lifetime of the universe.

[ Parent ]

probabilities (none / 0) (#90)
by speek on Thu Jul 18, 2002 at 09:47:21 PM EST

Yes, that's just how Penrose describes it in Emperor's New Mind - state space, and low entropy spaces are very small compared to higher entropy spaces, thus less likely that one would end up in one.

What you are saying is that the 2nd law is irreversible - or, maybe more precisely, the 2nd law is a statistical implication about an interesting property of a reversible system.

--
what would be cool, is if there was like a bat signal for tombuck - [ Parent ]

shit, typo (none / 0) (#123)
by speek on Fri Jul 19, 2002 at 06:35:16 PM EST

make that reversible, not irreversible.

--
what would be cool, is if there was like a bat signal for tombuck - [ Parent ]

Reversibility and entropy (5.00 / 1) (#106)
by tgibbs on Fri Jul 19, 2002 at 10:58:42 AM EST

Everybody who has tried to derive entropy from first principles has run up against the same problem: simply reverse the time axis, and  you have a world in which entropy decreases, and you haven't violated any physical laws (there are some subtle time assymetries in physics, but AFAIK, nobody has managed to tie them to entropy). So fundamentally, the Second Law has to arise from the boundary conditions. And for every set of boundary conditions that yields the Second Law, there is a time-reversed set that yields the opposite. And for "most" randomly selected sets of boundary conditions, entropy neither increases nor decreases with time (because it is already nearly maximal to begin with).

So to derive the Second Law, you have to dervive the boundary conditions for the universe. Basically, for entropy to increase with time, you have to start with submaximal entropy. So the Second Law leads ultimately into fundamental questions of cosmology. Interestingly, the Big Bang theory holds that shortly after the Bang, entropy was maximal (the universe was at equilibrium), and that our current state of relatively low entropy was generated by the expansion of the universe.


[ Parent ]

Boundary conditions (none / 0) (#120)
by bgalehouse on Fri Jul 19, 2002 at 05:53:11 PM EST

First of all, if the initial boundary conditions are simple, but not too simple, then there is nothing to keep my comments about RC4 excetera from applying globally. If the final boundary conditions are the same as the initial boundary conditions, things might start to reverse, but that doesn't change anything today.

Secondly, I'm not acquainted with whatever subtle time asymmetric laws you are referring to. However, do note that the entropy model works best when there is both a hot area and a cold area. If we isolate an area of deep space, it has so little happening that there is no change in entropy. Similarly, if we isolate a region of a star, there is so much happening that there is no change in entropy.

So the real question, is how these initial separations came about. Arguably it is the gradient which makes what we see possible. And so, we could enter the extremely speculative realm of big bang cosmology. However, without a unified field theory it seems an act of hubris to suggest that we are ready to understand this singular event which is so clearly both Quantum mechanical and Relativistic.

[ Parent ]

Re: 2nd law of thermodynamics (2.50 / 2) (#85)
by AzTex on Thu Jul 18, 2002 at 08:30:50 PM EST

speek, you said:

The second law is an observation made by scientists. There is no deeper understanding of the situation than that's what we see. People like Stephen Hawking talk about the 2nd law, not as though it's a law, but rather as though it's a bit of an anomaly. The 2nd law breaks with virtually all other laws of physics in that it is not reversible. Why does this one law seem to go only one direction in time?
The Second Law is reversible? How can a physical law be "reversible" or go a direction in time? What does that mean?

The way you throw these terms around, here and in your other comments, I suspect that you do not really understand what you are saying about any field of science. You are just regurgitating bits and pieces that you have read.

Thanks speek, but I'll stick with my understanding of thermodynamics in lieu of your "pop thermodynamics".



solipsism: I'm always here. But you sometimes go away.
** AzTex **

[ Parent ]
my understanding is imperfect for sure (4.00 / 1) (#122)
by speek on Fri Jul 19, 2002 at 06:32:34 PM EST

But, when you ask what reversible means, it really makes me wonder if you did read Wolfram's book, as he explains the concept quite well, and it makes me doubt you've ever even taken intro physics classes.

Please read all the rest of this thread, and learn - I am!

--
what would be cool, is if there was like a bat signal for tombuck - [ Parent ]

your understanding is nonexistant (none / 0) (#145)
by AzTex on Mon Jul 22, 2002 at 09:26:55 PM EST

speek, you are getting funnier every time:

But, when you ask what reversible means, it really makes me wonder if you did read Wolfram's book, as he explains the concept quite well, and it makes me doubt you've ever even taken intro physics classes.

I didn't ask what reversible means. I certainly already know!

I asked what you meant by saying that a physical law was reversible. And I ask what you meant by saying that a physical law goes a direction in time. Your statements to this effect are utter nonsense. The Second Law may have implications for our notions about the reversibility of a system and implications about the "arrow of time" but the Second Law does not get reversed or go in some direction.

So you again failed to answer my question.

Remember the first question I asked was for you to clarify your statement which was critical of my understanding of the Second Law in the comment which starts this thread. Remember, you started all this by saying:

the reviewer got stuck on his own misunderstanding of the second law of thermodynamics and went no further
I just wanted to know where I you think I have manifested some "misunderstanding". Come on. Tell me.

Funny!



solipsism: I'm always here. But you sometimes go away.
** AzTex **

[ Parent ]
funny man (none / 0) (#149)
by speek on Wed Jul 24, 2002 at 07:54:17 PM EST

You dismissed Wofram as a crackpot for suggesting the 2nd law of thermodynamics might not be universally valid. That strongly implies you have a flawed understanding of it.

Furthermore, you're comments about my discussion of reversibility is nothing but a strawman, as you deliberately choose to take my meaning incorrectly.

--
what would be cool, is if there was like a bat signal for tombuck - [ Parent ]

Re: worth reviewing, again... (none / 0) (#14)
by AzTex on Thu Jul 18, 2002 at 12:25:13 AM EST

Further clarification, sorry: I certainly did complete the chapter on Physics (Chapter 9). It's just that there was nothing in it but more hand waving. Wolfram runs simulations, waves his hands and says that they explain everything. Someone needs to fill in the "waves his hands" part for me.



solipsism: I'm always here. But you sometimes go away.
** AzTex **

[ Parent ]
he doesn't explain everything (none / 0) (#36)
by speek on Thu Jul 18, 2002 at 08:44:04 AM EST

He's pointing out that some interesting answers could come from this direction. He shows how a system of simple rules could account for things - he doesn't say it does.

--
what would be cool, is if there was like a bat signal for tombuck - [ Parent ]

Bah. (3.41 / 12) (#11)
by autonomous on Thu Jul 18, 2002 at 12:07:43 AM EST

This article is akin to slapping your grandfather upside the head when he makes a pass at a cute piece of ass because he still things he's a good looking man. Wolfram *IS* a genius. Wolfram *IS* a rich and well respected man. Wolfram *IS* an egocentric asshole like nearly every other academic. *NONE* of this should be surprising enough to warrant an article. This book is the result of his pet project, I have a pet project too, and like Wolfram, if I ever write a book about my pet project, I don't think I'd peer review it either. My peers don't get to comment on my pets, be they projects or women.
-- Always remember you are nothing more than a collection of complementary chemicals worth not more than $5.00
As well (none / 0) (#12)
by autonomous on Thu Jul 18, 2002 at 12:09:05 AM EST

If I ever write a book about my pet project, I will spell check it before I press post.
-- Always remember you are nothing more than a collection of complementary chemicals worth not more than $5.00
[ Parent ]
Wolfram is smart, but he has not been.. (3.33 / 3) (#15)
by Weezul on Thu Jul 18, 2002 at 12:44:29 AM EST

..connected to the rest of the academic community for a long time.  There are informed smart people saying simillar things, but they don't seem to be taking Wolfram too seriously.  The bottom line is that being smart is not everything in academics.. and Wolfram seems to be missing most of the rest of it.

As a side note, most scientific acadmeics I know are pretty humble.  If your in the sciences for any significant length of time, you quickly discover that there is a hell of a lot you don't understand.  I don't think Wolfram ever got past that "I've defeated every problem I've tackled" phase.  Now, one might start to think that all liberal arts academics were egotistical from reading the stuf that find its way into the popular press, but I have a fealing that is partialy the presses fault.

"Fascism should more appropriately be called Corporatism because it is a merger of state and corporate power." - Benito Mussolini
[ Parent ]

Perhaps (none / 0) (#16)
by autonomous on Thu Jul 18, 2002 at 12:47:58 AM EST

Perhaps he didn't get a chance to "get over" that stage because he was busy getting his PHd when everyone else his age was primarily concerned with getting laid and passing their highschool algebra course. Its hard to be humble when you're brilliant. I'm not saying he's got merit, I'm not saying you need to listen to him, I'm just saying he's earned the right to stroke his ego and pet project without the crushing criticism.
-- Always remember you are nothing more than a collection of complementary chemicals worth not more than $5.00
[ Parent ]
I strongly disagree (5.00 / 2) (#23)
by JetJaguar on Thu Jul 18, 2002 at 02:28:08 AM EST

I'm just saying he's earned the right to stroke his ego and pet project without the crushing criticism.

I have to strongly disagree with this statement. I don't care what his credentials are. When he presents a pretty unconventional set of theories that have nearly zero experimental support, and he doesn't come up with the support himself, the work deserves to be looked at with a lot of skepticism. The fact that he got his PhD at a young age, is arguably a "genius," and spent x number of years working on this is totally irrelevent. Crap is still crap no matter how long you spend trying to dress it up as something else.

Nobody, no matter how smart they are, has earned the right to peddle any scientific theory based only on their credentials alone. Some strong evidence must be provided or the whole work definitely merits strong criticism.

If science worked the way you are advocating, nearly every field of science would be turned upside down. There are a good number of formerly great scientists that contributed great discoveries in their fields, who later in their careers have advocated some really nut-ball theories. Chip Arp is just one of several examples in astronomy.

[ Parent ]

*sigh* (none / 0) (#24)
by autonomous on Thu Jul 18, 2002 at 02:46:18 AM EST

I'm not advocating that he is peddling anything scientific. Its is NOT a scientific paper, it is NOT peer reviewed, it IS his personal pet project. Get over it. It is a book, it may be more or less important than a cheap paperback romance novel. Let the man have his pet project.
-- Always remember you are nothing more than a collection of complementary chemicals worth not more than $5.00
[ Parent ]
double sigh (none / 0) (#62)
by JetJaguar on Thu Jul 18, 2002 at 01:21:03 PM EST

It doesn't matter what audience Wolfram's book is addressed to. I also don't understand how you can say that he isn't peddling anything scientific, especially when "A New Kind of Science," is right in the title of the book. It sure sounds like he's peddling something scientific to me, and doing so in a very irresponsible fashion. To say the least, many of the claims he's making probably should have been submitted for peer review before he wrote the book. There are two reasons for this:

1. If Wolfram is right, he makes another very significant contribution to the scientific process, and many more scientists will begin (almost immediately) to build on his work.

2. If Wolfram is wrong, the scientific community can work out the bugs before getting the general public all riled up over this only to find out later it was all a bunch of crap.... Remember cold fusion?

By doing an end run around the peer review process, or not making it abundantly clear that a lot of his conjectures are little more than speculation, he does a great disservice to science, and to his readers. And in the end ultimately hinders the scientific process by muddying the waters and making it more difficult for his peers to evaluate his work.

By the way, there's nothing wrong with having a little pet project, there's even nothing wrong with telling people about it, but I have to draw the line at the point he starts to claim that everything can be modeled by CA with no other reason than it simply happens to be his pet project.

[ Parent ]

whatever (none / 0) (#66)
by autonomous on Thu Jul 18, 2002 at 03:00:14 PM EST

Nobody has ever needed credentials to create a book and publish it. EVER.

Wolfram is publishing a book on his pet project, it has a title that implies a scientific slant, but it is NOT peer reviewed, we've established it is NOT properly referencing or referenced, it is just a book. The book is a brain dump of what Wolfram has been toying with for years, if it was scientific, it would be a paper, but it isn't, so don't treat it like it should be.

If your going to get upset, you should perhaps start by addressing the fact the new age and religion section in every (nonspecialty) bookstore in the world is 100x larger than the science section. You should get upset that half the books in the science section are about as far from science as you can get. You should get upset that nobody except me is getting upset about that. You should not get upset that Wolfram has written a book.

He may or may not be a crackpot, but he's got a hobby and its keeping him from flooding academic journals with utter crap like many of his (should be retired) peers are. Don't get upset because you think it should be science, its a book, published by a man who wants people to read his pet theory.

As for cold fusion, I believe that was announced in a scientific journal wasn't it? Its been a while and I don't recall. Besides, who the hell cares what the public thinks, most of them still think there is someone sitting on a cloud watching over them, they'll get worked up with or without Wolframs help. It would appear however, the only worked up people are the scientists who think every publication should be (or is) science.
-- Always remember you are nothing more than a collection of complementary chemicals worth not more than $5.00
[ Parent ]
Peer review is rarely done on physics books. (5.00 / 1) (#103)
by claudius on Fri Jul 19, 2002 at 09:30:49 AM EST

Disclaimer: I am a professional physicist; all my comments here are based upon my experiences as a physicist.  (Side note to streetlawyer: I am probably at least as humble as most lawyers I know).

I think you overstate the importance of "peer review" of books.  Unlike the humanities, physicists (and scientists in general) tend to read and write journal articles rather than books.  A piece of work that appears in a refereed journal article carries much more clout than would the same work showing up in a book.

Books on physics, by and large, are not peer reviewed.  Most scientists writing a book, however, are more than willing to give a draft copy to as many of their colleagues as they can find who are willing to proofread the text.  This is to avoid possible embarrassment from mistakes, omissions, or leaving out a key reference.  Often editors and publishers will encourage a careful technical proofreading of a text, but this is not mandatory for getting a book into print.  It is merely a business decision on the part of the publisher: texts that generate favorable reviews sell more copies; accurate, well reasoned physics books receive more favorable reviews.

Because of the less exacting standards of book publication in the sciences, many scientists who opt to write the book do so for specific reasons.  Perhaps they wish to teach from a textbook that more accurately mirrors their own perspective on a given subject, perhaps they have some ideas they'd like to develop that require more space or more license than even review-article journals would allow. Perhaps they just like seeing a big rag-bound tome with their name on it.  Seldom, however, do they do it to "get the word out" on their work.  Journals are just so much more efficient, timely, and effective for information dissemination.

I don't fault Wolfram for not going through a peer review process with his book.  I do fault him for writing a pretentious, self-aggrandizing pile of sludge, for appropriating the work of others without proper attribution, for suing a former employee of his for trying to publish in a scientific journal work of his that Wolfram perceived as "stealing thunder" from ANKOS, and for Wolfram being a titanic asshole the few times I've had the misfortune to deal with him in person.

[ Parent ]

I'm an astronomer... (none / 0) (#141)
by JetJaguar on Mon Jul 22, 2002 at 01:14:07 AM EST

I'm also (until recently) a professional astronomer, and I am aware of how things work with regard to publishing books. One big difference I see here is that, normally, when someone writes a science book, the peer-review has already been done. I can't think of a single legitimate science book that covers material that has not already gone through some form of formal peer-review.

I agree that usually it's not necessary for most physics books to go through a formal peer review process simply because nearly all of the material presented there has already gone through peer-review. It would be silly to force someone that's writing a layman's book on Maxwell's equations to go through formal peer-review. It's just not needed.

However, it does not look like a lot of what Wolfram has written has gone through that process, and to me at least that is another mark against it. Of course, it is also quite likely that Wolfram has not put his work up for review for many of the reasons you state, it wouldn't surprise in the least.

[ Parent ]

humble (3.00 / 1) (#34)
by streetlawyer on Thu Jul 18, 2002 at 07:27:42 AM EST

As a side note, most scientific acadmeics I know are pretty humble

Most scientific academics I know have damn good reason to be a lot humbler. Very few physicists, in particular, are even half as intelligent as their behaviour might justify.

--
Just because things have been nonergodic so far, doesn't mean that they'll be nonergodic forever
[ Parent ]

Peer review is less a concern than proper citation (4.00 / 2) (#42)
by claudius on Thu Jul 18, 2002 at 09:28:35 AM EST

Wolfram made a strategic blunder in not referencing his work properly; this mistake has cost him credibility and cold cash in lost sales.  Every scholaly pursuit that I know of has general guidelines for how one goes about appropriating the work of others.  Wolfram has a history of ignoring the reference conventions used in science and mathematics, instead reporting the work of others without attribution (or with only vague, 3-point-font attribution at the end of a book chapter), a practice that would be called plagiarism if the person doing the appropriating were not a MacArthur Fellow.  He was criticized repeatedly for this back when he was a "real" scientist, and the behavior is evinced even more dramatically with his magnum opus.  This puts many people off from the get-go, and it diminishes his credibility both with the scientist and, as pointed out by the reviewer, with the layperson.  

Of course, who am I to criticize someone who has this to say about himself: "Stephen Wolfram is the creator of Mathematica and is widely regarded as the most important innovator in scientific and technical computing today.  A distinguished scientist and award winning writer, he is the founder and president of Wolfram Research--the world's leading technical software company."  -- The Mathematica Book (3rd edition) by Stephen Wolfram (Cambridge University Press, New York, 1996).  

[ Parent ]

it really is a book review (4.50 / 4) (#17)
by khallow on Thu Jul 18, 2002 at 12:49:16 AM EST

While I've seen some complaints that he hasn't "read" the book, I need to remind some people that you don't need to read an entire book cover to cover to figure out that you hate it. AzTex has struggled through 450 pages of an almost 1200 page book. That sounds reasonable to me.

Disproving the second law of thermodynamics with a computer model sounds pretty flaky to me too.

While this story appears destined for the bit bucket, I want to point out a couple of things. First, this is the first review of this book that I've read that is pretty unbiased. The author is neither prostituting for the book nor defending the ivory tower of Science from the uncouth barbarians. And has been pointed out before, this is the first true review on K5 of "A New Kind of Science".

Stating the obvious since 1969.

it's unfortunate (none / 0) (#54)
by speek on Thu Jul 18, 2002 at 11:26:44 AM EST

The reviewer actually seemed clearly biased to me. If the review gave you the impression that Wolfram thinks he "disproved the second law of thermodynamics with a computer model", then it's shitty review. Please read some of the other comments that give a better account of the book.

--
what would be cool, is if there was like a bat signal for tombuck - [ Parent ]

Early opinion (none / 0) (#72)
by forii on Thu Jul 18, 2002 at 04:43:14 PM EST

I'm still trying to work my way through the book. Not because the material is tough, as I have studied cellular automata and artificial life, and have been familiar with Wolfram's work for years now. But instead, I am finding it more and more difficult to read his hand-waving explanations, and grand "scientific" conclusions drawn from what are mostly aesthetic judgements.

I'm mostly trying to power through it, on the hope that maybe there is something redeeming further ahead in the book. But by now I'm not especially hopeful.


Proud member of the ACLU, the NRA, and the EFF.
[ Parent ]

but is it a review (5.00 / 1) (#104)
by tgibbs on Fri Jul 19, 2002 at 10:25:08 AM EST

While I've seen some complaints that he hasn't "read" the book, I need to remind some people that you don't need to read an entire book cover to cover to figure out that you hate it. AzTex has struggled through 450 pages of an almost 1200 page book. That sounds reasonable to me.
If you want to say, "I started the book and couldn't finish it; I hated it," that's fine, but you probably shouldn't claim to have written a "review".

[ Parent ]
Enough Wolfram Already.... (2.00 / 4) (#18)
by thelizman on Thu Jul 18, 2002 at 01:00:49 AM EST

Besides the fact that he gets disproportionate coverage here, someone has already done a review of Wolframs book.
--

"Our language is sufficiently clumsy enough to allow us to believe foolish things." - George Orwell
where? (4.80 / 5) (#55)
by speek on Thu Jul 18, 2002 at 11:34:10 AM EST

Yes, let's talk more about Israel and the Palestinians, please? Or how about Bush and Cheney, they haven't been mentioned in a few minutes. MPAA, RIAA anyone? What do you think about copyright?

Wolfram is clearly squeezing out all these more worthy discussions.

--
what would be cool, is if there was like a bat signal for tombuck - [ Parent ]

2nd law of thermodynamics (3.00 / 1) (#20)
by xriso on Thu Jul 18, 2002 at 01:12:37 AM EST

Wolfram was creating non-natural systems, right? Why should we expect natural laws to apply to his systems?
--
*** Quits: xriso:#kuro5hin (Forever)
Pardon? (none / 0) (#21)
by war crimes on Thu Jul 18, 2002 at 01:16:26 AM EST

The second law of thermodynamics applies to closed, non-natural systems.

Frankly, I am not convinced that the distinction between natural and non-natural systems even makes sense.

----
UID 33259 is trhurler's troll account.
[ Parent ]

Re: pardon (none / 0) (#83)
by tgibbs on Thu Jul 18, 2002 at 07:42:48 PM EST

The second law of thermodynamics applies to closed, non-natural systems.
No, it is quite possible to imagine systems very different from the natural world in which the second law does not apply. It is even possible to simulate such systems on a computer in a world where the second law does apply

[ Parent ]
Got up to page 504 myself and just quit reading (4.50 / 8) (#22)
by keenan on Thu Jul 18, 2002 at 01:36:49 AM EST

By this point he began trying to create a model of the universe that were filled with supposition upon supposition that were based on absolutely nothing that had been concretely observed in reality.  Being a math major (with a keen interest in physics), I can understand the need to approximate reality with simplistic devices that don't necessarily apply perfectly, but he simply goes too far.  In his arrogance, he seems to repeatedly attempt to alter reality to fit his model, as opposed to altering the model to fit reality which is the standard scientific method.  Every time he said something like "And if the universe is described by a multiway system..." he was losing me more and more -- he never had any good evidence to show that the model really applied.  

But the final straw was that he repeatedly talked about properties of the universe that can be inferred if it were a cellular automata, but these properties have no measurable effect on reality.  For example, he talks about "the sequencing of events in the universe" in Chapter 9, "demonstrating" that when using a cellular automata to represent the universe that our concept of time does not require the universe to have a global clock.  There is no purpose of this investigation:  we have absolutely no knowledge of whether the universe uses a global clock and there's no way we could know through our observations alone.  We could hypothesize using Occam's razor, but it's a question akin to what occurred before the big bang -- it provides no predictive power.  In science, you simply don't propose such an unprovable theorem:  science is about making hypotheses that can be proven or disproven with experimentation.  He simply does away with this -- I really don't see how he can call it "A New Kind of Science" when it's not really science whatsoever.

Keenan

Relativity (5.00 / 3) (#31)
by greenrd on Thu Jul 18, 2002 at 07:18:22 AM EST

For example, he talks about "the sequencing of events in the universe" in Chapter 9, "demonstrating" that when using a cellular automata to represent the universe that our concept of time does not require the universe to have a global clock. There is no purpose of this investigation: we have absolutely no knowledge of whether the universe uses a global clock and there's no way we could know through our observations alone.

According to General Relativity, the universe does not in fact have a global clock. An event that to an observer on a spacecraft takes t seconds may be measured as taking t+1 seconds by an observer elsewhere. Such time dilation effects can become very pronounced under special circumstances, such as near to the event horizon of a black hole. Also, even without black holes, one can generate a so-called "twin paradox" (which is not really a paradox at all, just a surprising outcome), by flying very close to the speed of light relative to the Earth, which causes you to age slower than people back on earth.

Furthermore, these time dilation effects (though in this case they are miniscule, fractions of a millisecond) have been observed to occur precisely as General Relativity predicts, by experiments with aircraft. Therefore, in this sense the theory that the universe does not have a global clock has in fact been validated by experiment.

Perhaps, however, you were talking about something slightly different. If so, please elaborate.


"Capitalism is the absurd belief that the worst of men, for the worst of reasons, will somehow work for the benefit of us all." -- John Maynard Keynes
[ Parent ]

The twin paradox (none / 0) (#94)
by Gumpzilla on Fri Jul 19, 2002 at 12:08:58 AM EST

The twin paradox is not that you age less when you're traveling at relativistic velocities. The paradox comes in when you consider, that, from the reference frame of the person on the spacecraft, it appears that the person who remains at home is traveling at relativistic speeds. So the person on the craft could argue that everybody on the earth, which is traveling at relativistic speeds relative to his craft, would experience time dilation and age less than him. Yet when he returns, he finds this is not the case. Thus the paradox.

There's a good discussion of this in Griffith's book Introduction to Electrodynamics, a pretty common undergraduate E&M text. I'd imagine some Googling could turn up good stuff as well.

[ Parent ]
Resolved (5.00 / 1) (#117)
by Khedak on Fri Jul 19, 2002 at 01:56:19 PM EST

So the person on the craft could argue that everybody on the earth, which is traveling at relativistic speeds relative to his craft, would experience time dilation and age less than him. Yet when he returns, he finds this is not the case. Thus the paradox.

I am just an amateur in this field, but I'm pretty sure that this paradox is resolved by observing that the dilation is a function of acceleration, and not relative velocity. That is, the person on earth has not undergone a series of extreme accelerations, wheras the person on the rocketship has. This is what resolves the paradox, and the person who underwent massive accelerations is the person for whom times "slows down." The same happens at the event horizon of a black hole (theoretically):

The person accelerating away from the universe into the black hole sees the universe speed up ad inifinitum until he or she is destroyed. An outside observer sees the hapless black hole victim slowing down as he accelerates down the gravity well, until he or she becomes unobservable.

[ Parent ]
Pretty much correct, as I understand it. (none / 0) (#127)
by Gumpzilla on Fri Jul 19, 2002 at 08:45:03 PM EST

The conventional answer is, as you mention, that the two frames can be distinguished because one of them experiences acceleration. I was just pointing out that the twin paradox is not just differential aging from time dilation, it's the "symmetry" argument that I mentioned.

[ Parent ]
really? (4.50 / 2) (#33)
by streetlawyer on Thu Jul 18, 2002 at 07:24:29 AM EST

we have absolutely no knowledge of whether the universe uses a global clock

Are you claiming to have an argument against Einstein as well as Wolfram? One of the things that general relativity does tell us is that there is no single temporal ordering, let alone a global clock.

--
Just because things have been nonergodic so far, doesn't mean that they'll be nonergodic forever
[ Parent ]

Causality (none / 0) (#96)
by majubma on Fri Jul 19, 2002 at 01:33:20 AM EST

But that in and of itself does not mean that there is no global clock or "privileged" reference frame; just that there seem to be any number of such reference frames that work just as well.


--Thaddeus Q. Thaddelonium, the most crookedest octopus lawyer in the West.
[ Parent ]
I think you misunderstood (5.00 / 2) (#53)
by speek on Thu Jul 18, 2002 at 11:23:58 AM EST

Wolfram intentionally goes through a lot of types of systems as possible models of the universe (such as multiway systems). He pares down the list of possible systems to A) match with current experimental results (in other words, he's NOT ignoring what's currently known), and B) to make the search easier. So he explains his thought processing as he tries simple systems and discards them as inappropriate, or unlikely.

The global clock example is a good instance of Wolfram removing unnecessary assumptions from his models. Most of his simple rule systems do, in fact, require a global clock, so it was important for him to show that there were ways these rules could work without one.

He does do a lot of conjecturing. I see nothing wrong with that. He's not proving anything - he's trying to indicate some directions that might yield fruitful research.

--
what would be cool, is if there was like a bat signal for tombuck - [ Parent ]

Concrete reality? (none / 0) (#82)
by tgibbs on Thu Jul 18, 2002 at 07:38:34 PM EST

By this point he began trying to create a model of the universe that were filled with supposition upon supposition that were based on absolutely nothing that had been concretely observed in reality.
I'd be curious to see what aspects of, say, string theory you believe to have been concretely observed in reality. He seems to be doing pretty much what other cosmological theorists do: asking whether a particular abstract model can generate experimentally observed phenomena, such as special relativity.
we have absolutely no knowledge of whether the universe uses a global clock and there's no way we could know through our observations alone.
However, there is a purpose if you are trying to construct such a model of the universe. At this point, he is not asking whether the universe has a global clock. He is asking whether a model of the sort he is proposing requires a global clock to emulate certain observed characteristics of the universe.

[ Parent ]
Permutation City by Greg Egan (4.20 / 5) (#25)
by IvyMike on Thu Jul 18, 2002 at 03:23:48 AM EST

Spoilers ahead...

I'm working my way through Wolfram's book, but I'm constantly being reminded of the science fiction book Permutation City by Greg Egan. Even though it was published in 1995, it reads like fiction someone would write AFTER having digested Wolfram's book.

The book revolves around (and takes place in) something Egan calls the "Autoverse", which is a cellular automaton large enough to hold an entire simulated universe. And in fact, near the end of the book, (and here's the spoiler) the Autoverse has intelligent residents. These creatures, who are advanced enough to be looking for the Grand Unified Theory of their universe, are told that their universe is in fact a giant cellular automaton. They reject that hypothesis as implausible.

In any case, as I read "A New Kind of Science," I keep wondering if I'm also having difficulty with Wolfram's work because it's too implausible. ;)



This will sound sarcastic (4.00 / 1) (#35)
by FredBloggs on Thu Jul 18, 2002 at 08:41:16 AM EST

and i really dont mean it to be, but how many sci-fi books are about people taking life seriously only for the reader (if not the inhabitants in the book) to discover that its all just a big game? I get the feeling its more than just this one!

[ Parent ]
Not this one... (none / 0) (#61)
by dark on Thu Jul 18, 2002 at 01:20:32 PM EST

Permutation City is entirely different. It is _not_ about people suddenly realizing they live in the Matrix. I detest spoilers so I won't tell you anything more about it :)

[ Parent ]
Reviews (2.00 / 1) (#29)
by Master Of Ninja on Thu Jul 18, 2002 at 06:38:12 AM EST

I thought about getting this book, but there were a few things that [put me off. First of all the price - 40 RRP!! I thought that was a bit steep for a book. Then there are the reviews. I know that really I should read it first and then judge, but the negative reviews coupled with the high price just put me off buying the thing. I'll probably read it when (if?) the library gets a copy.

Here's the links to the amazon reviews - you can make up your own minds:
Amazon UK
Amazon US

yeah... (3.00 / 2) (#60)
by chopper on Thu Jul 18, 2002 at 12:38:43 PM EST

i was put off that it cost me like 60 bux.

then i realized, that's about a dollar a pound, which isn't too bad at all.

give a man a fish,he'll eat for a day

give a man religion and he'll starve to death while praying for a fish
[ Parent ]

Reading vs Skimming (3.50 / 2) (#30)
by AzTex on Thu Jul 18, 2002 at 07:02:32 AM EST

There seems to be some concern in these comments (especially the editorial ones) that I didn't actually read the book. This would be a valid concern if it were true.

I carefully read Chapters 1-9 (pp 1-546) and the Notes for Chapters 1-9. I skimmed Chapters 10-12 (pp 547-848) and the Notes for Chapters 10-12.

The Notes (pp 849-1200) are quite extensive. I probably should have mentioned in my review that they were in many cases more interesting than the corresponding chapters.



solipsism: I'm always here. But you sometimes go away.
** AzTex **

Poor review (-1) (4.94 / 19) (#32)
by Kalani on Thu Jul 18, 2002 at 07:18:37 AM EST

First of all, you should have put this in edit mode.  What are "cellular automations?"  For somebody who's read so much about them, I'd expect that you'd spell the name correctly.

Now I'll just try to respond point by point:

Reading the Preface and Chapter One, you may notice that Wolfram makes extensive use of the word "I" and of the phrase "this book". In these sections, he also takes the opportunity time and time again to tell us just how long it has taken him to write "this book" and how much effort it has taken him to write "this book". Should we be impressed with the author for his labor?

He uses the word "I" because the book (before the notes section) is an informal text for the layman.  I suspect that he inserted the information about how long it took him to get to finishing the book (and so on) because he might have been concerned that some people would look at the systems he analyzes early on as trivialities.  You could have simply read the notes and seen for yourself that he's collected and attempted to synthesize a lot of information.

However, work in this field has been done before.

The second item in the list returned by google for your search is an analysis of Conway's "Game of Life."  The "Game of Life" is a 2D cellular automaton that uses two colors and nearest neighbor rules.  Conway (a mathematician at Cambridge in the 60s, I think) issued a challenge at one point to try to find unbounded growth in his system.  He was surprised when it was discovered (the so-called "glider gun") because he'd originally thought that no such growth would ever occur.  A short time after this discovery, it became evident that the "Game of Life" could emulate the behavior of a universal computer by certain constructions manipulating streams of gliders (where gliders are the sort of bits in the system).  This capability has actually been proven ... as the second link in your list most likely will tell you.  In Wolfram's book, you'll find a demonstration of the universality of the 2-color 1D Rule 110 Cellular Automaton (proof by Matthew Cook -- a former employee of Wolfram Research).  This demonstrates that universality is possible in cellular automata systems that are even simpler than the "Game of Life" (further, Wolfram goes on to actually analyze the whole class of CA systems into which the "Game of Life" falls).  The demonstration of this fact about Rule 110 is interesting in and of itself.  It is achieved by noting that Rule 110 immitates a cyclic tag system (which is known to support universal behavior).

Also, while pointing out that "work in this field has been done before" (as if that invalidates his current work), you might also mention that the very system of ordering the 1D CA ("Wolfram Numbers") originally came from Wolfram himself.  While the layman text doesn't include a history lesson, there's plenty of information in the notes and on google to see where Wolfram fits into the history of CA research.

Frankly, I was left wondering what is Wolfram's original work and what has been done before.

I think that this is a valid criticism.  It's unfortunate that he chose to only refer to previous work in small type (in the preface and in the notes).  Worst of all is that Cook gets little more than a passing mention despite the fact that the universality of Rule 110 demonstrates one of Wolfram's theses (that computationally irreducible complexity has a low "underlying rule complexity" threshold) very well.  Wolfram was actually engaged in a legal dispute with Mr. Cook when he tried to publish his results before Wolfram was ready to release the book.  I'm surprised you didn't mention that.  Why didn't you?

Wolfram uses the outcomes of the experiments of earlier chapters an an attempt to relate simple algorithms to natural phenomena. But most of it is hand-waving.

What did you feel was hand-waving, in particular?  The turbulent fluid flow and manifold gas examples are, I think, very convincing.  They're actually used currently by researchers.  There are actually more natural phenomena to which CA models are being applied than Wolfram covers in the book.  So there's not really a lot of dispute over whether or not they're useful in certain problems (more on that later).

He appears to make some headway with crystals and simple examples of plant growth. He shows photographs of snowflakes, leaves and stems and compares them somewhat favorably with the patterns produced by particular cellular automations. There are, of course, no exact matches.

OK, so they don't look like exact matches.  They're clearly very close.  It's near enough to be useful (just like other near-fit tools of analysis applied to problems -- "particles" and forces and so on).  You also are kind of taking that whole section out of context if you're implying that he fails to put together a successful "theory of everything" here.  In that chapter he was only trying to show that CA are useful as modelling tools.  And it's true, they are.  Take a look at the example of the snowflakes and you'll see that the algorithm he applies generates a whole series of snowflakes that demonstrate certain patterns at each step (arms, edges, thickness, etc) and that each one of these generated patterns has a matching actual snowflake (pages 370-371).

Continuing with biological examples, Wolfram shows a number of photographs of common seashell and animal shapes. Curiously, he does not attempt to compare these to any cellular automaton or other results.

This is incorrect.  He shows how patterns like the ones seen on animal coats or on the surface of conch shells is very very similar to particular cellular automatons (the pattern on the surface of the shells, for instance, is the most commonly found 2-color 1D CA in the book and I'm surprised that you wouldn't even remark on it).

Also he demonstrates an algorithm that will generate every possible conch shell!  The method he uses is similar to the 2D substitution system he uses earlier.  Did you not mention this simply because it wasn't a cellular automaton?  The cyclic tag system is strictly-speaking not a cellular automaton either, yet it is used in the proof of the universality of the Rule 110 automaton.  Wolfram demonstrates the connections between all of these different kinds of systems.  Doesn't that seem at least significant enough to mention?

Over the course of discussion of real biological structures and pigmentation patterns, Wolfram makes it clear he does thinks that natural selection does not explain the observed patterns of biological systems. But he does not actually state why he holds this opinion. He says that he thinks that cellular automata algorithms are being chosen at random and natural selection has nothing to do with it.

Completely untrue!  Wolfram states that he believes that natural selection plays a secondary role in that it deselects systems of needless complexity.  That's actually a key point.  He makes his opinion on all of this very clear.  Like with the example of the procedure that generates every possible conch shell, there is a similar procedure to generate, for instance, every possible cheetah coat.  Natural selection comes into play when the cheetahs with the conspicuous coats get killed off (or whatever other reason might cause a particular iteration of the coat-generator to not show up in nature).  So clearly natural selection still explains, in a certain sense, why you see the patterns that you see.  The actual algorithms for a given species of animal, for instance, are "known" beforehand -- and the differences come from different members of the result set being used (or the algorithm iterating more/less times).

By that same token, natural selection also serves the purpose of getting rid of algorithms that are too random (class 3) or too simple (class 1).  I don't think that's such a crazy idea.  It might not be correct (I don't know for sure), but it's not the sort of absurd notion that you paint it as.

At this point I don't judge the remainder of your review to be worth analyzing.

You might want to look more closely at his "Principle of Computational Equivalence" and how it relates to all of the other systems you neglected to mention (but which Wolfram describes in explicit detail).  You might also want to find out what some well-known researchers who've worked in the same area have to say and try to read the arguments on all sides by people who've been analyzing these systems for a long time.  You might get more out of it than the idea that the embodiment of ideas in this area can be summed up as "the Universe is a computer."

-----
"Nothing says 'final boss' like a giant brain in a tube."
-- Udderdude on flipcode.com

thank you (5.00 / 4) (#49)
by GoStone on Thu Jul 18, 2002 at 10:44:02 AM EST

I would really appreciate a proper review of this book by someone who understands the issues. Your 'comment' is worth more than the poor article it is commenting on.


Cut first, ask questions later
[ Parent ]
How do you distinguish those who understand. . . (none / 0) (#95)
by Gumpzilla on Fri Jul 19, 2002 at 12:36:16 AM EST

I thought the original article presented some fairly well reasoned, coherent objections to the book. The parent of your comment raises objections to those objections by saying that the author of the article was either choosing to misrepresent Wolfram or did not adequately understand Wolfram. How do you know it's not the other way around? Though the author of the parent comment (I'm starting to write like Wolfram now, how apropos) does have some textual evidence to back him up, it's certainly not much, and not really anything on this "Principle of Computational Equivalence."

After that comment, I'll throw in my own two cents. I think this book has stirred up a lot of emotional, knee-jerk reactions on both sides. Lots of people I know who consider themselves scientific dismiss it as crackpot science without having read it, saying that it exhibits so many dubious qualities that it's not worth the investment in time. Other people here seem to be taking essentially the opposite tack. I've started reading the book to decide for myself, and it's pretty tedious. I can't really recommend it.

The first six chapters feel like an Encyclopedia of Simple Systems, in some sense, and get fairly boring to read quite quickly. Read a few pages, see a few pages of pictures demonstrating that system X exhibits complex behavior, move on to system Y. I read the first four chapters and then skimmed five and six. I then started reading the seventh chapter, hoping for interesting insight rather than further cataloging of simple systems. After getting about 25-30 pages in, I hit a block. Wolfram's lack of useful citation finally got to me. He makes a claim suggesting that he has evidence that at least some natural processes produce randomness through some intrinsic mechanism instead of randomness from random initial conditions. One argument for this is that intrinsically produced randomness could be repeatable if the initial conditions are tightly monitored enough. Now, to quote:

So has such repeatibilty actually been seen in practice? Unfortunately there is so far very little good information on this point . . . . But scattered around the scientific literature . . . I have managed to find at least some cases where multiple runs of the same carefully controlled experiment are reported, and in which there are clear hints of repeatability even in behavior that looks quite random.
(pg.326)

Nothing in the main text, at least in this chapter, elaborates further on what these cases might be. Checking the notes does not yield particular articles but "temperature differences in thermal convection in closed cells of liquid helium, reactions rates in oxidation of carbon monoxide on catalytic surfaces . . . ." Wolfram is making it difficult for other people to go look at this stuff for themselves. This does not inspire confidence in me; if I had experimental support for a point I wanted to make, I'd wave it around in the air for all to see. Why make people jump through hoops if it's saying what you want it to say anyway?

I think other posters might have the right idea. Read the book as new science, with scientific expectations, and you're likely to be frustrated and disappointed at best. Read the book as exposure to interesting ideas, and not worry too much about what's fresh and original, and you might enjoy it.

[ Parent ]
Egos a-gogo (none / 0) (#100)
by GoStone on Fri Jul 19, 2002 at 04:47:47 AM EST

Simply this. I'm interested in Cellular Automata and similar ideas. Now someone has written a popular book on the subject, which seems to have a lot of original ideas, mixed in with less original stuff (presumably like a text-book plus research article combined).

Every review I have seen so far focusses on the issue of Wolfram's ego, which I am not very interested in, massive though it may be. Sure it may be a fault in the book, but I'd rather reviewers focussed on the interesting bits.

Of course I am on shaky ground because I haven't read the book yet, but still I found the comment at the root of this thread contained more interesting ideas than the article itself, and I suspect the book is more interesting than the ego-watchers are letting on. Negative criticism is always easy. The greater thing is to illuminate the positives.


Cut first, ask questions later
[ Parent ]

I already posted a link to a .. (none / 0) (#102)
by DeHans on Fri Jul 19, 2002 at 06:19:12 AM EST

review collection, but these reviews are high on technical details:

A general scientific look: Review by Brian hayes, American Scientist
Focusing on QM: Book review by Scott Aaronson (PDF), Quantum Information and Computation
A mathematical look: Review by Henry Cohn, Mathematical Association of America.

Look at the collection page for more reviews.

[ Parent ]
Wolfram (3.00 / 1) (#70)
by jmzero on Thu Jul 18, 2002 at 04:15:45 PM EST

I've read a few reviews of the book, though admittedly not the book itself.  Usually, it seems, the objections to the book center around "Well, it's not like he invented all of this, why does he think he's so great?"

The book would indeed have to be earth shattering to satisfy those familiar with its hype.
.
"Let's not stir that bag of worms." - my lovely wife
[ Parent ]

Not again... (none / 0) (#84)
by merkri on Thu Jul 18, 2002 at 08:28:50 PM EST

Frankly, I was left wondering what is Wolfram's original work and what has been done before.

I think that this is a valid criticism.

Actually, I believe this is the most important criticism that can be leveled against this book, or any book for that matter.

Wolfram's book, from what I've read through, seems to have a lot of great content. However, Wolfram seems to confuse integrating and adding to existing knowledge with creating it.

This, in my mind, is just as important as what the actual content of the book is. If nothing else, it amounts to intellectual plagiarism and is inexcusable. However, it also obfuscates his arguments by making his line of reasoning less clear, and by omitting reference points of knowledge, so to speak, that might lead to refutation of his claims.

You condone his frequent use of the first person, and his boasting of what a tremendous accomplishment the text is, as conveying to the "layperson" that his examples aren't trivial. This interpretation is hard to accept, especially given the poor or nonexistent citations of existing work. And if his book truly reflects "a new kind of science"--if it really is that important--he should aim it to technical audiences as well as lay audiences. In fact, he should aim it more at technical audiences than lay audiences, and document exactly what he is speaking of when he speaks of it, so as to make his reasoning (and any possible gaps) more clear.

It's all to easy to confuse a book that summarizes existing knowledge (even if the author himself contributed it) with one that adds to it in some way. This especially true if the author has taken lengths to obscure the difference, as I believe Wolfram has done.

[ Parent ]

I disagree (none / 0) (#87)
by Kalani on Thu Jul 18, 2002 at 09:28:43 PM EST

Actually, I believe this is the most important criticism that can be leveled against this book, or any book for that matter.
That's certainly true if you're a politician, but I think that most people are simply interested in whether or not the book contains any interesting/useful ideas. Nobody got on Newton's ass when he wrote his first law in the Principia without adding "* - as originally postulated by Galileo Galilei at the start of the 17th century," nor did they throw out the Principia when Hooke charged Newton with plagiary. In fact, to quote from the best biography of Isaac Newton:

"... Newton had seemed to invite objections, albeit in a rather backhanded way. Let it be said on Hooke's behalf that he couched his correction in the mildest of terms ... Nevertheless, he did express his conviction that under conditions of no resistance a body let fall on a rotating earth would not fall to the center but would rather follow forever a path resembling an ellipse. [Hooke] explicitly treated the problem as orbital motion by referring to it as "my theory of circular motions compounded by a direct motion and an attractive one to a center."
Yet somehow people found it in their hearts to acknowledge the interesting observations made by Newton in the Principia. Things like this are an even more glaring fault in Newton's Principia than in Wolfram's layman text specifically because the Principia was intended for a professional audience. You'll find some references in Wolfram's notes section (probably not as many as most people would like, but then I suspect that Wolfram has also duplicated many of the small results that will cause the most fuss from theorists with less far-reaching ideas -- to get an idea for how somebody with an equal amount of time/effort in this field feels about Wolfram, see Ed Frenkin's comment above).

But if you're going to deride him for a lack of references in the popularization (although he does still mention the big ones: Conway, Turing, von Neumann, Goedel, etc) then you'll also have to deride many other books about cutting-edge theory directed at a layman audience (Brian Greene's "The Elegant Universe," Stephen Hawking's "A Brief History of Time," and even Newton's "Optiks").

Anyway, on to the rest of your comment ...

Wolfram's book, from what I've read through, seems to have a lot of great content. However, Wolfram seems to confuse integrating and adding to existing knowledge with creating it.
Look at the notes section and you'll see a pretty good history of what he's done and where he fits into it all. He didn't come up with the idea of cellular automata, that belongs to von Neumann. He did do much in the area of reducing their complexity to simplest forms and systematizing the whole classification scheme (Wolfram numbers, etc). You make it sound like he's a bit player stealing the great insights from other people. He has created a lot of existing knowledge, and he's also integrated many other types of systems into a coherant framework (everything from combinators to Turing machines!). That's not at all insignificant.

It's all to easy to confuse a book that summarizes existing knowledge (even if the author himself contributed it) with one that adds to it in some way. This especially true if the author has taken lengths to obscure the difference, as I believe Wolfram has done.
It might be that you don't know the difference, but he didn't set out to write a book that explicitly delimited every minute ownership of a fraction of an idea. There IS, of course, a place for that (you'll find enough information in the notes to be able to determine how progress had been made in the field up to now and who's been making it). His book is clearly NOT just a summary of existing knowledge. That's like saying that Newton simply summarized the existing knowledge of the method of infinitesimals and quadratures originally outlined in Wallis's "Arithmetica Infinitorium," or that Maxwell simply summarized existing knowledge of electricity and magnetism.

It seems like many who "review" the book only do so with a superficial knowledge of its contents. That explains all of the strawmen arguments.

-----
"Nothing says 'final boss' like a giant brain in a tube."
-- Udderdude on flipcode.com
[ Parent ]
I disagree (none / 0) (#93)
by merkri on Thu Jul 18, 2002 at 11:58:36 PM EST

That's certainly true if you're a politician, but I think that most people are simply interested in whether or not the book contains any interesting/useful ideas.

I disagree, at least in a certain sense. Textbooks contain lots of interesting ideas--at least good ones--but there's usually nothing substantially new or revolutionary in them. People will like good textbooks because they contain interesting ideas, but that doesn't necessary mean that we should treat the author of an excellent textbook like the next Newton just because they can summarize a field well.

My feeling reading A New Kind of Science that it's much like reading a corrupted but excellent sort of textbook, a bastardized introduction to CA. Wolfram does an excellent job of introducing and summarizing CA and its relevance to other fields. My problem with him is that, unlike a textbook, he fails to reference these things with any sort of care whatsoever, and in fact goes further, often explicitly or implicitly attributing things to himself which clearly should be attributed to others. Just as in a good textbook, there's excellent discussion of the current state of things, but little that is truly new. Wolfram desparately wants it to be new--his--so he couches it in language that tries to make it seem so.

Comparisons to Newton I think only contribute to Wolfram's hype. Putting aside any issues of whether or not Newton did plagiarize (and is therefore not due all the credit usually attributed to him), one might argue that Newton did propose testable predictions, theories to be tested. Wolfram, in my opinion, does not do this. Like much of CA work, he points out how much such and such a pattern resembles some natural feature without explaining how the CA can lead to specific predictions about real outcomes. In fact, he sometimes implies that such predictions cannot be made--leading to the vague and unhelpful conclusion that some thing looks like something else, but cannot truly be used for predicting it with any degree of accuracy. Perhaps this is true, but like much else in the book, it is certainly not new.

Many have criticized Wolfram for confusing models with reality, and this is relevant to how much Wolfram is going out on a limb in this text. I'm not sure that anyone did say that complex patterns could not be generated by simple processes. Is this true? Maybe I'm wrong, but I thought that was suggested in some form centuries ago by philosophers such as Hume; it forms the basis for theories and models of chaos and self-organization, which have been discussed for years and integrated into numerous other fields, such as evolutionary biology (c.f. Kaufmann, 1993). Wolfram does an excellent job of delineating the details from one perspective, but credit for that belongs to a field, not Wolfram; to attribute to him the premise that the complexity of nature can be attributed to simple causes and processes is a bit too much.

You make it sound like he's a bit player stealing the great insights from other people. He has created a lot of existing knowledge, and he's also integrated many other types of systems into a coherant framework (everything from combinators to Turing machines!). That's not at all insignificant.

I don't mean to imply that Wolfram hasn't contributed immensely to CA. What I do mean is that the field he describes and the thesis he develops in A New Kind of Science are not solely his doing. He may be a major contributor to CA theory, but that doesn't mean that we should attribute to him major concepts like the idea that simple processes account for the complexity of nature.

The problem I have with the text is that I have yet to hear anyone--especially those in the field--say that there is anything both new and revolutionary in A New Kind of Science. I do not hear controversy over a "crazy idea" that may or may not work, I hear controversy over whether or not some ideas are his or not. Usually the answer is "no, that's not his idea, but it's cool, isn't it?" The text doesn't come across to me like the unusual ravings of a crackpot or genius, but rather, like the familiar arrogance of someone who wants us to believe that decades of work in a field should be attributed to him. Wolfram may be a major contributor, but a major contributor does not make a creator.

I guess I may change my mind when someone convinces me that there is something in the text that is radically new and represents a "new kind of science." Until that point, my assessment of the text is that it is an excellent summary of an interesting field, written by an author who attributes too much of it, implicitly or explicitly, to himself.

[ Parent ]

OK, so you don't know what you're talking about (none / 0) (#98)
by Kalani on Fri Jul 19, 2002 at 02:31:03 AM EST

My feeling reading A New Kind of Science that it's much like reading a corrupted but excellent sort of textbook, a bastardized introduction to CA.
If you actually think that his book is just a textbook on CA, why did you even respond to me to begin with? You only picked out the "he hasn't referenced everybody in the field" comment and didn't pay any attention to anything else. If you'd actually read the book you'd know that it wasn't about CA in particular. I won't respond again if you haven't got anything to say about the actual content of the book.

often explicitly or implicitly attributing things to himself which clearly should be attributed to others
I'm just curious which results you're thinking of in particular? Could you point them out?

Wolfram desparately wants it to be new--his--so he couches it in language that tries to make it seem so.
Yet you freely admit that you aren't familiar with the area of study and you haven't cited any specific instances of that by him (except for the obvious numerous instances of his stressing the importance of the work in the layman version of the text). I think that, with your comments on what he desperately wants, you're more willing to examine the psychology of the author than the content of his book.

Comparisons to Newton I think only contribute to Wolfram's hype.
It wasn't a comparison to Newton, it was a demonstration of precedent, or are you willing to say that such behavior was allowed for Newton and not anybody else? It hasn't yet been established that Wolfram's insights are on the scale of the introduction of Newton's fluxions and comprehensive physics. Obviously he shouldn't just vault into the annals of history without fair skepticism and investigation. What you propose is not anything even approaching that.

one might argue that Newton did propose testable predictions, theories to be tested. Wolfram, in my opinion, does not do this
Then you simply haven't read the book. Wolfram applies his systems to various problems in nature as well as other realms (meta-mathematics, for instance). Why don't you go back and read the section on the relationship between mathematical deduction and multiway systems, demonstrating an intuitive explanation for Goedel's Theorem?

Besides, model-fitting is primarily about coming up with the best description of what you already know (usually empirically) to be true, and then going from there to find out if the most precise model you find makes any new predictions. Wolfram doesn't claim to have a complete model of the whole universe ... he claims to have systematized the study of systems like CA to such a degree that they're fit for more formal analysis and application to many different fields. So I'm afraid that your "opinion" doesn't have much justification.

Like much of CA work, he points out how much such and such a pattern resembles some natural feature without explaining how the CA can lead to specific predictions about real outcomes.
Read the book, you don't know what you're talking about. He clearly shows in the example of, for instance, conch shells that every configuration can be generated by an iterative 3D substitution system. He describes how certain other such simple systems can describe how various features of natural systems emerge. He also goes over his preliminary work in applying such systems to other areas. Just read the book.

Many have criticized Wolfram for confusing models with reality, and this is relevant to how much Wolfram is going out on a limb in this text.
So some people have no idea what they're talking about. That's not surprising to me at all. Many people seem unwilling to actually read the book and consider it. If they had, they'd know that such a suggestion is idiotic because he hasn't done anything of the sort.

I'm not sure that anyone did say that complex patterns could not be generated by simple processes.
That's part of the popular account. If you actually read the book (meaning the notes section in particular) you will obviously see that his thesis is NOT "complex patterns can be generated by simple processes!" That's been known by mathematicians long before any "philosopher" gave any input.

Perhaps when you've actually read the book you might try commenting on its contents. I'll be willing to respond again when that is the case.

-----
"Nothing says 'final boss' like a giant brain in a tube."
-- Udderdude on flipcode.com
[ Parent ]
OK, whatever makes you feel better about yourself (1.00 / 1) (#116)
by merkri on Fri Jul 19, 2002 at 01:23:21 PM EST

If you actually think that his book is just a textbook on CA, why did you even respond to me to begin with?...If you'd actually read the book you'd know that it wasn't about CA in particular.

You are grossly mischaracterizing my position, presumably to enhance your own. I didn't say it was actually a textbook, but rather, that it read to me like a corrupted textbook. If Wolfram had limited the scope to CAs, and not tried to make the text about "a new kind of science", it might have been a good text. As it is, he didn't, and so we have to judge it that way. In any event, good textbooks aren't always limited in scope to their immediate topic, but discuss applications and extensions as well.

I won't respond again if you haven't got anything to say about the actual content of the book.

In your posts, you seem to have a habit of dismissing people's comments as so beneath yourself that they're not worthy of commentary. If they truly weren't, you simply wouldn't. The fact that you do, but say that you shouldn't, exemplifies the same sort of arrogance that characterizes Wolfram's text (e.g., "If you don't appreciate the genius of this text, you just don't understand it and are not worthy of reviewing it or having any commentary whatsoever." It's silly, and the prevailing attitude I've seen from those who are favorable to the book.).

I'm just curious which results you're thinking of in particular? Could you point them out?

How many pages and topics are there in the book? In  my opinion, in many cases in which Wolfram goes outside the topic of CA, he exaggerates the novelty of what he says. Often, it's not so much that he always says he did something when he didn't, but that he implies as much or obscures the fact that he didn't. My reaction to some of the content regarding random numbers was very much that way, for example (I think he mischaracterizes pseudorandom number generation somewhat, or at least obscures certain characteristics, although his discussion was still interesting); I also had similar reactions to his discussion of the development and evolution of biological systems (c.f. the book by Kaufmann I cited earlier). And in, any case, when I said I'm not in the field, I meant the field of CA. I do work in the field of statistics and bioinformatics, though. The way that Wolfram characterizes his work, it's difficult for anyone to "not be in the field".

Anyway, you yourself provide an example of Wolfram's intellectual egocentrism with Mr. Cook:

Wolfram was actually engaged in a legal dispute with Mr. Cook when he tried to publish his results before Wolfram was ready to release the book.

It's not as if Wolfram denies Mr. Cook's contribution, just that he goes to some lengths to obscure it (and no, a brief footnote in the back of the text is not sufficient for exactly this reason).

It wasn't a comparison to Newton

It wasn't? I must have misread your original post.

are you willing to say that such behavior was allowed for Newton and not anybody else?

I'm not even saying it should have been allowed for Newton. But Wolfram isn't Newton anyway.

It hasn't yet been established that Wolfram's insights are on the scale of the introduction of Newton's fluxions and comprehensive physics. Obviously he shouldn't just vault into the annals of history without fair skepticism and investigation. What you propose is not anything even approaching that.

I guess this gets to the heart of what irritates me so much about this book. First, in what you wrote above, it is implicitly assumed that these are indeed "Wolfram's insights". Why is this? Because he says or implies so? That very assumption is highly questionable, if for no other reason that Wolfram expands the relevance of the content to almost every intellectual domain imaginable. Much of the criticism, in my opinion, hasn't been fair enough: in most cases, it is the responsibility of the author to justify a position; Wolfram takes the position, at least implicitly, that such tasks are the responsibility of the readership. Second, even if they were all his, you're right: it hasn't been shown that these observations are on the scale of anything Newton, Einstein, Godel, or anyone else ever wrote. I suspect they may never be. The reason why is because there isn't anything new or "risky" in the text.

For example, if I were to say to someone, go and show that that there is something in the text that is (a) Wolfram's original idea, and (b) predicts something that is radically different from what has already been known, what would they do? First, they'd have a difficult time showing that Wolfram does indeed present anything new. Even if they did, however, I suspect they'd have a difficult time finding anything that poses a novel and useful prediction to be falsified. It's successfully making risky predictions that represents progress in science (at least in the Popperian view), and I don't see any of that in A New Kind of Science. Maybe Wolfram means to imply that in his "new kind of science" it's enough to make comments about vague similarities?

Read the book, you don't know what you're talking about. He clearly shows in the example of, for instance, conch shells that every configuration can be generated by an iterative 3D substitution system. He describes how certain other such simple systems can describe how various features of natural systems emerge. He also goes over his preliminary work in applying such systems to other areas. Just read the book.

I have read the book. And the similarities are interesting. But I recall very similar comparisons being made between fractals and natural features years ago. When I say that the models are vague, what I mean is that they generally lack a certain degree of quantitative predictive precision, a certain precise correspondence between elements of reality and the elements of his models. They are models, true, but in the text, things are generally limited to Wolfram presenting a snapshot of a CA aside some natural feature, and saying "Look how similar they are!". I don't mean to imply CAs aren't useful--they are tremendously useful--just that they don't represent "a new kind of science".

And in any event, none of this is really relevant to my major complaint about the book, which is that the important ideas in the text are not Wolfram's.

His thesis is NOT "complex patterns can be generated by simple processes!"

Then what is his thesis? What would you characterize as the thesis of the book? That's the popular account because that's what the thesis of the text is. Perhaps it's not phrased as Wolfram would have it be phrased, but that's it in essence. Maybe something about computational irreducibility? That's hardly revolutionary either.

That's been known by mathematicians long before any "philosopher" gave any input.

My point about the book exactly. And in any event, the line between math and philosophy is thin and often arbitrary.

Anyway, I do like the book on some level, and think there are wonderful observations that are made within. Some of those observations do actually derive from Wolfram's earlier work. But the text, to me, reads more like the chronicle and documentation of a field--much like a textbook--than the presentation of a bold new thesis. If Wolfram had recognized that fact, and treated it as such, I might have respected the text, and him, much more. But he tries to expand the relevance of the book and novelty of his ideas to just about everything; in the process, as a reader, I am forced to shift my evaluation of the text from being about CA to larger concepts about complexity and predictability. And at that point, he fails to convince. If he had only written about CA, or if he had been more honest and realistic about the novelty of his observations in general, I would have a different opinion of the text. But as it is, he didn't, and so I don't.


[ Parent ]

For those looking for more reviews, ... (4.80 / 5) (#39)
by DeHans on Thu Jul 18, 2002 at 09:06:00 AM EST

or more in-depth reviews, look at a Collection of Reviews of Wolfram's A New Kind of Science

I have to agree with this review (4.00 / 1) (#41)
by DesiredUsername on Thu Jul 18, 2002 at 09:26:26 AM EST

I slogged through the numerous CA examples and finally hit what I thought was the point of the book--the applications. In the biology section it was obvious he had no idea what he was talking about (or did, but was waving his hands around). I had to give up part way through the physics chapter because either he was talking way above my head (which was pointless to read) or he was wrong (which was going to confuse my attempts to actually learn physics).

Play 囲碁
hand waving (4.00 / 1) (#58)
by speek on Thu Jul 18, 2002 at 11:44:45 AM EST

To some extent, I agree he does hand-waving in biology to distract you from wondering how CA's can be applied usefully in biology. My brother is a bio-chemist, and in talking with him, it really seems that Wolfram's approach is quite similar to the way most biologists approach things. Biologists are not dominated by calculus-based mathematics the way physics is, and they would love to be able to model biological entities with computer programs. About the only thing Wolfram differs in is that I think he would recommend trying to "idealize" the situation and create extremely simplistic computer models. Trying to hard to match all the data may be a mistake, in fact. Also, I think Wolfram is suggesting that certain evolutionary sequences are likely to happen again and again and again, not because they are the most advantageous traits to develop, but because the underlying rules are the same, and always tend to develop in those directions.

As far as physics goes, I think his approach differs more dramatically with the current approach. Take the 3-body problem, for example. Wolfram suggests, it'd be easier to solve if you used a CA-type system, rather than a constraints based system. He also suggests that chaos theory is not the right explanation for the difficulty - and to some extent this is testable (he describes how one might test it in the book).

And, the idea that the universe might consist of a network of nodes is pretty elegant - and not an unheard of suggestion.

--
what would be cool, is if there was like a bat signal for tombuck - [ Parent ]

2nd Law of Thermodynamics (4.00 / 2) (#43)
by beardy on Thu Jul 18, 2002 at 10:22:32 AM EST

I couldn't help but wonder about this story when I found the following article on the BBC website
which says that the second law is reversible at very small distances and timescales according to an experiment by two australian researchers.

Call it fate, call it karma. It just seemed to fit with this entry.

---------------------------------------- "...and then he injected me, twice!!!!!"
thank goodness (2.44 / 9) (#44)
by turmeric on Thu Jul 18, 2002 at 10:25:29 AM EST

society is not emotionally ready for cellular automata based science. it is simply too powerful. we would only use it to make new weapons and kill each other rather than helping each other.

i am glad wolfram has printed this crackpot book. hopefully this will set back CA research a few decades due to the 'crackpot' shadow cast over it.

PS. there have been algorithms to make plant-looking things for many many years, they are all over the computer graphics discussion communities even old videos like 'the minds eye' show results. hell even any of the earl 90s books on computer fractal graphics are full of this sort of thing. alan turing even said in the mid 20th century that he had almost worked out daisies or dandelions or some such flower.

(computer graphics, well, before that quake 3d bullshit came a long, that is, and now people only care about texture mapping blood stains and body parts blasting apart)

anyways i am glad this research will be stopped .... humanity is emotionally immature and fully developed CA might lead to advances that would render this planet uninhabitable, might make nuclear-threat-cold-war seem like a fun time at the carnival. imagine 'made to order' biochemical weapons that are built out of extremely simple parts that simply replicate themselves. like, for example, a wind-borne seed (ie the way that weeds propagate) ... but it releases toxic nerve gas or something. i am sure the army would love this, but i would not love it at all.

This guy sounds unamerican. (1.10 / 10) (#46)
by TIPS on Thu Jul 18, 2002 at 10:30:20 AM EST

This crackpot thinking "outside the box". Hrm. Probably a communist too I bet, just like that other one. Probably a Jew too.

e pluribus oculus! - - TIPS is for kids!
Comparison to Einstein is bogus (3.66 / 3) (#80)
by jbuck on Thu Jul 18, 2002 at 07:09:32 PM EST

Einstein was well-versed in the physics of his day, and published his work in peer-reviewed journals. He referenced the work of other researchers and used their terms, e.g. Lorenz contraction, not Einstein contraction. He produced a theory that made testable predictions, and these predictions were in fact tested and shown to be correct. Wolfram, on the other hand, used his millions to self-publish a vague piece of work that has no references, makes no testable predictions (as far as I know, please correct if I am wrong), and that received no peer review. Crack. Pot.

[ Parent ]
There is now experimental evidence of this point.. (4.00 / 1) (#47)
by GoStone on Thu Jul 18, 2002 at 10:32:46 AM EST

By coincidence in the latest edition of New Scientist there is an article about this very point. An australian team have demonstrated that the 2nd law does not apply to small systems. On the micro-scale not the nano-scale I think.


Cut first, ask questions later
posted in the wrong place, sorry (none / 0) (#48)
by GoStone on Thu Jul 18, 2002 at 10:35:21 AM EST

meant to be in the discussion thread by speek


Cut first, ask questions later
[ Parent ]
it's ok (none / 0) (#56)
by speek on Thu Jul 18, 2002 at 11:35:37 AM EST

I see it :-)

--
what would be cool, is if there was like a bat signal for tombuck - [ Parent ]

Well.. (none / 0) (#92)
by mindstrm on Thu Jul 18, 2002 at 11:02:30 PM EST

upon first read...

The experiment they did was not a closed system. It was taking input from outside (shaking).

[ Parent ]

Bad Timing and the Second Law (5.00 / 2) (#59)
by ZahrGnosis on Thu Jul 18, 2002 at 12:13:32 PM EST

Personally I think the Second Law of Thermodynamics is flawed, but that's just me. The BBC, however, just published this story today about it.

the "second law" is observed globally... (5.00 / 2) (#73)
by demi on Thu Jul 18, 2002 at 04:53:37 PM EST

but not always locally. There are numerous examples of reverse entropic chemical reactions and processes in nature, but there is always an overall increase in entropy if you take the surroundings into consideration. It just depends on how entropy is thermally quantified in each case, which is not trivial.

Personally I think the Second Law of Thermodynamics is flawed, but that's just me.

What is its flaw, and what leads you to believe there is one to begin with?

[ Parent ]

I'll bite (none / 0) (#108)
by kubalaa on Fri Jul 19, 2002 at 11:20:40 AM EST

First, a "law" which is only observed in a statistical sense isn't much good. It's like the weatherman issuing the "20% chance of rain law". The problem being there's no such thing as a closed system.

Second, non-conservation is somehow repugnant. Not a good objection, but hey. It's the age-old problem; why do things change? Why are things different?

[ Parent ]

Don't be silly (none / 0) (#110)
by ghjm on Fri Jul 19, 2002 at 11:45:41 AM EST

You might as well argue that conservation of energy isn't much good because it is also only meaningful in closed systems. Suppose you are lying on a tropical beach and observe that the sand around you is growing hotter, with no decrease in energy anywhere else around you. Conservation of energy has been violated! Except of course that the energy came from the sun.

The fact that the only closed system is the universe means that the second law must be understood as flows of entropy from hither to yon, just as with conservation of energy. It's nothing more complicated than that.

-Graham

[ Parent ]

except (none / 0) (#121)
by speek on Fri Jul 19, 2002 at 06:12:06 PM EST

It's not entirely certain that the universe is a closed system.

--
what would be cool, is if there was like a bat signal for tombuck - [ Parent ]

flows of entropy (none / 0) (#143)
by kubalaa on Mon Jul 22, 2002 at 10:20:36 AM EST

That's just the point -- there is no conservation of entropy. The second law states that, in a closed system, entropy will appear out of nowhere! This feels untidy, like there's something wrong with our concept of entropy.

[ Parent ]
Wow, I'm convinced (none / 0) (#105)
by Fon2d2 on Fri Jul 19, 2002 at 10:34:51 AM EST

Now that I read that incredibly vague article by the BBC about the violation of the 2nd Law, I'm convinced. The 2nd Law of Thermodynamics must be flawed.

[ Parent ]
2nd Law of Thermodynamics (5.00 / 1) (#136)
by mumble on Sat Jul 20, 2002 at 10:33:52 AM EST

The second law of thermodynamics just states: a closed system will tend to occupy the configuration with the highest number of equivalent states. The probability of being in a particular state is 1/N where N is the total number of possible states. The second law comes in, because highly ordered configurations have only very small numbers of "equivalent" states, while highly random configurations have a huge number of "equivalent" states.
(Tip: think of a normal distribution which has a really high peak, and very short tails on either side. The highly ordered configuration is off in the wings, while the higly random configuration is the centre of the distribution.)

So being in a highly ordered configuration may have probability of i/N (where i -> 1 as N -> inf), the probability of being in a highly random state has probability of m/N (where m -> N as N -> inf). In plain English, by the sheer numbers involved, (eg. Avogadro's number = 6*10^23) the probability of being in a highly random configuration is ~ 1, and the probability of being in a highly ordered configuration is ~ 0. That is the 2nd law of thermodynamics (I wish I could explain it simpler though, because it really is a simple concept).

The reason the results in the BBC and New Scientist article are essentially meaningless is because the system is so damn small!! 100 beads!!! Compared to ~10^23. So instead of a highly ordered configuration having probability of 1/N = 10^-23, we have 1/N = 1/100. So run that experiment about 100 times, and it is almost certain you will see a "violation" of the second law. Albeit in a totally trivial way, IMHO.
(Note you can include the time factor as increasing the effective size of N. The longer you observe it, the larger the effective N is. Hence tying in with them saying it only violated the system if they kept the time interval of observation short.)



-----
stats for a better tomorrow
bitcoin: 1GsfkeggHSqbcVGS3GSJnwaCu6FYwF73fR
"They must know I'm here. The half and half jug is missing" - MDC.
"I've grown weary of googling the solutions to my many problems" - MDC.
[ Parent ]

my two cents (4.00 / 2) (#68)
by bukvich on Thu Jul 18, 2002 at 03:44:59 PM EST

I have purchased and seen every page of _New Kind of Science_. I have read each word on about the first 400 pages. For what it is worth . . .

1.) the pictures are fabulous. The pictures on page 377 are quasi-orgasmic.

2.) it would be a much better book if it was competently edited. It is repetitive.

3.) for a "scientific" book it is a very fast read, largely because it is so repetitive.

4.) it is more of a publication event than a scientific event. I would compare it to Godel-Escher-Bach and Emperor's New Mind. Both in significance and in quality. I enjoyed reading those other books immensely. When I have finished, I am pretty sure I will have enjoyed reading this book immensely.

5.) Cellular Automaton experts (I am not one) can love it or hate it or whatnot. To me their opinions of this book are not data points.

That is all.

B.

I rate it well below Godel, Escher, Bach (none / 0) (#115)
by Humuhumunukunukuapuaa on Fri Jul 19, 2002 at 12:56:51 PM EST

Because GEB was extremely enteraining as well as having a pretty damn rigorous (for a popular exposition) proof of Godel's Theorem. The stuff on AI is dated now but so much of the rest of GEB was wonderful.

Both Wolfram and Penrose, on the other hand, are weak by comparison. They both have brilliant minds but they have too many crackpot ideas that don't stand up to peer review and so chose to publish in popular science instead.
--
(&()*&^#@!!&_($&)!&$(*#$(!$&_(!$*&&!$@
[ Parent ]

Peer-review (none / 0) (#119)
by tgibbs on Fri Jul 19, 2002 at 05:11:19 PM EST

Both Wolfram and Penrose, on the other hand, are weak by comparison. They both have brilliant minds but they have too many crackpot ideas that don't stand up to peer review and so chose to publish in popular science instead.

I don't think that it is true that Wolfram's ideas have failed to stand up to peer review. So far as I know, he has not attempted to publish his recent work in peer-reviewed journals.

I'm not sure that a peer-reviewed journal is really all that appropriate for a work of this nature. Peer-review works best for papers that make straightforward and well-defined additions to an accepted body of research, where it is easy to identify experts who are competent to judge the work. Wolfram's stuff doesn't really fit into this category. From a practical point of view, Wolfram doesn't really need peer-review, because he has sufficient credibility based upon his previous accomplishments to attract an audience without the "seal of approval" of peer review. And since he is not an academic, he is doesn't need to add peer-reviewed publications to his c.v.

There are a lot of intriguing ideas in the book. The real question is whether they lead anywhere useful. Some may remember the excitement over Thom's "Catastropy Theory" some years back, yet very little in the way of real scientific accomplishment seem to came out of it. I think that Wolfram's book will probably stimulate research in the area, which is likely Wolfram's goal.

[ Parent ]

Interesting to compare with Catastrophe Theory (none / 0) (#129)
by Humuhumunukunukuapuaa on Fri Jul 19, 2002 at 09:29:15 PM EST

I'm part way through a book on the subject at the moment.  There's nothing wrong with the subject - it's all interesting and the mathematics is rigorous and insightful.  I guess I'm not reading all the hype that went with it so I'm not suffering from any kind of disillusionment.  I can't understand why it's died.  (Actually I think maybe it hasn't - the mathematics now lives on but has been recategorized to the fields where it should have been all along - somewhere in a backwater of differential topology).


Wolfram doesn't really need peer-review, because he has sufficient credibility based upon his previous accomplishments

Scientists always need peer review.  The best and smartest scientists all have pet theories that are outside of their domain of knowledge meaning that they need interaction with scientists in those other fields to moderate their work.  Wolfram is most definitely stepping well beyond the bounds of his knowledge in ANKOS.  What Wolfram is good at is CAs.  In ANKOS CAs are only the starting point from which he jumps to all kinds of speculation.
--
(&()*&^#@!!&_($&)!&$(*#$(!$&_(!$*&&!$@
[ Parent ]
Oy (none / 0) (#133)
by Kalani on Fri Jul 19, 2002 at 10:12:07 PM EST

Scientists always need peer review.
Scientific theories need peer review. For that matter, mathematical proofs need peer review (though a proof is true whether or not eyes have seen it, of course). That's the stage that all of this is at right now.

Wolfram is most definitely stepping well beyond the bounds of his knowledge in ANKOS. What Wolfram is good at is CAs.
Tell that to the guys at CalTech who gave him a PhD in physics. You could also see this comment I've referred to in another post which comes from Ed Fredkin (see other posts for why he's significant) adressing a claim similar to the one you make above.

In ANKOS CAs are only the starting point from which he jumps to all kinds of speculation.
He actually analyzes how universality and computationally irreducible complexity apply to a wide variety of systems, not just CA. The layman part of the book starts out with CA because they're the simplest systems he analyzes (and that's also what makes the proof of the universality of Wolfram's 1D CA #110 interesting).

Although there is a great deal of speculation (which is generally more believable when you read the more detailed version in the notes), there is also much beyond it. After the 9th chapter he comes out of the mess of speculation in many different areas to explain the discoveries that prompted him to speculate in those areas in the first place.

-----
"Nothing says 'final boss' like a giant brain in a tube."
-- Udderdude on flipcode.com
[ Parent ]
Catastrophy Theory (none / 0) (#137)
by tgibbs on Sat Jul 20, 2002 at 03:07:27 PM EST

There's nothing wrong with the subject - it's all interesting and the mathematics is rigorous and insightful.
Yes, this is the point. There doesn't seem to be anything wrong with it, but it didn't seem to lead much of anywhere. Sometimes an idea seems really interesting and intriguing, but it neverthess fails to yield any major advances. That doesn't mean that it's wrong. It may just be that the knowledge or understanding to take it further simply doesn't exist at the time. Conversely, sometimes an idea that is completely wrong results in major advances, because it spurs other work that yields significant knowledge. A lot of Wolfram's ideas strike me as cool in much the same way that I thought Catastrophe Theory was cool, so I'm wondering whether they will turn out to be a fertile ideas or sterile ones.
Scientists always need peer review. The best and smartest scientists all have pet theories that are outside of their domain of knowledge meaning that they need interaction with scientists in those other fields to moderate their work.
Yes, and I don't doubt that Wolfram interacted with other scientists in writing his book--there is a huge list of acknowledgements--but that isn't really the function of peer review. Peer review functions as the "gatekeepers" to publication in major journals, so that readers know in advance that the author of a paper was able to convince at least a couple of informed scientists that what he was saying had value. On the other hand, there's a long list of scientists whose work I would happily read even without peer review, because I trust their work far more than the opinion of an anonymous reviewer.

[ Parent ]
Peer-reviewed Wolfram (none / 0) (#126)
by Kalani on Fri Jul 19, 2002 at 08:23:42 PM EST

You might be interested in taking a look at Cellular Automata and Complexity, a collection of Wolfram's earlier papers on cellular automata. They're really very interesting.

As for this current work, there's clearly a lot of speculation in it. However, if his Principle of Computational Equivalence is valid then it represents a synthesis between systems like CA, multiway systems, register machines, (essentially all of the various systems he describes in the book) and actual complex systems in the real world. I personally think that the idea that I'm equivalent to weather patterns in a computational sense is interesting (it's also similar to ideas that other people have had, but Wolfram does actually have a more specific description of what this means and how it plays out than I've ever seen). One of the more interesting ideas that comes out of Wolfram's principle is that Goedel's theorem arises as an essentially obvious characteristic of many different kinds of complex systems. He demonstrates this by showing how mathematics (that is, the practice of deriving theorems from axioms and so on) is analogous to a multiway system (even to the degree that multiway systems have a comparable definition of logical consistency) and how certain questions about a multiway system that goes above some threshold of complexity are equivalent to the question Goedel asked, and so incompleteness is a sign of having reached a particular level of complexity (which is suggestive of universal computation and so on). That's where the interesting ideas are in this book (although I really enjoyed the chapter on "perception and analysis" just for the clarity with which certain concepts in data compression and encryption were presented). I think that all of the comments on the "theory of everything" (which is clearly incomplete and which, I think, Wolfram should have excluded from the book) are red herrings.

-----
"Nothing says 'final boss' like a giant brain in a tube."
-- Udderdude on flipcode.com
[ Parent ]
What do you mean... (none / 0) (#130)
by Humuhumunukunukuapuaa on Fri Jul 19, 2002 at 09:34:55 PM EST

.../his/ Principle of Computational Equivalence.  Isn't it a completely obvious hypothesis that's been made many times.  I can remember dozens of ...er...stoned...conversations as a student

"hey man!  maybe the vortices in this cup are some hyperintelligent being we can't understand"
"shut up and finish making the coffee".

The same goes for the applications Godel's Theorem.  Penrose himself makes similar suggestions but the idea goes back many years before that.

--
(&()*&^#@!!&_($&)!&$(*#$(!$&_(!$*&&!$@
[ Parent ]
Well here is where Wolfram cites prior work! (none / 0) (#131)
by Kalani on Fri Jul 19, 2002 at 09:55:59 PM EST

If you listen to the NPR "Talk of the Nation" interview with Wolfram here, you'll see that me makes the remark, "people have often said that the weather has a mind of its own."

Anyway, Wolfram actually has proven this property of multiway systems. As for the "Principle of Computational Equivalence," he presents much more evidence for it (and defines it more rigorously provided that you're happy with his complexity classification scheme) than any of your stoner friends have (but if that's not the case please direct me to their lengthy expositions of this principle).

-----
"Nothing says 'final boss' like a giant brain in a tube."
-- Udderdude on flipcode.com
[ Parent ]
Also (none / 0) (#132)
by Kalani on Fri Jul 19, 2002 at 10:01:45 PM EST

Another difference between Wolfram and your stoned friends is that he's demonstrated how certain axioms or certain rules in particular kinds of rule-based systems will generate sufficient complexity to support universality and has suggested why it happens and how often it happens, while your stoned friends just shut up and finished making the coffee.

-----
"Nothing says 'final boss' like a giant brain in a tube."
-- Udderdude on flipcode.com
[ Parent ]
I heard (1.00 / 1) (#139)
by trane on Sun Jul 21, 2002 at 02:59:45 AM EST

Wolfram shoots heroin.

[ Parent ]
He certainly sounds like a smack junkie (none / 0) (#140)
by Kalani on Sun Jul 21, 2002 at 05:10:21 PM EST

You, on the other hand, still need to lay off the crack pipe!

-----
"Nothing says 'final boss' like a giant brain in a tube."
-- Udderdude on flipcode.com
[ Parent ]
My whole point was... (none / 0) (#142)
by trane on Mon Jul 22, 2002 at 07:18:39 AM EST

your comments as to "stoners" added an annoying layer of "ad hominem" to your otherwise logically-based line of reasoning...but I probably shouldn't have posted. if this was a wiki, I could remove it!

[ Parent ]
You're being oversensitive (none / 0) (#144)
by Kalani on Mon Jul 22, 2002 at 07:59:23 PM EST

I have nothing against stoners. The point was only that there's a lot more detail to Wolfram's principle than "what if the vortices in my cup of coffee are like a hyperintelligent being?"

There are plenty of intelligent people who've done drugs. The validity of an idea has nothing to do with whether or not the originator of the idea has done drugs.

-----
"Nothing says 'final boss' like a giant brain in a tube."
-- Udderdude on flipcode.com
[ Parent ]
Can I just point out (none / 0) (#78)
by Mental Blank on Thu Jul 18, 2002 at 06:17:01 PM EST

That the plural of cellular automaton is cellular automata. :)

evolution & entropy (none / 0) (#81)
by tgibbs on Thu Jul 18, 2002 at 07:26:32 PM EST

Over the course of discussion of real biological structures and pigmentation patterns, Wolfram makes it clear he does thinks that natural selection does not explain the observed patterns of biological systems. But he does not actually state why he holds this opinion. He says that he thinks that cellular automata algorithms are being chosen at random and natural selection has nothing to do with it.

From the tone of Wolfram's discussion of the role of natural selection in biology, he obviously is not well versed in either.

I don't find anything particularly wrong with Wolfram's understanding here. The idea that not every characteristic of an organism needs to be adaptive is well accepted in evolutionary biology. Wolfram makes a pretty good case that the fine details of these patterns may not reflect selective optimization. It may be that selection simply dictates the broad characteristics of the pattern and the details are irrelevant. By showing that there is a broad class of algorithms that can generate similar patters, Wolfram makes a good plausibility argument that the fine details of such patters might be fortuitous rather than adaptive.

Then based on the supposed violation by this model and some other irrelevant discussion of order and randomness, he states on page 457 his Earth-shattering conclusion that The Second Law of Thermodynamics is not valid.
...for particular classes of cellular automata. He is not making an argument about the validity of the second law in the physical world, other than suggesting that if, as he speculates, some simple algorithm describes the universe, the 2nd Law could turn out to be only approximately valid.

Dumb Question (none / 0) (#86)
by R343L on Thu Jul 18, 2002 at 08:31:24 PM EST

Could someone explain clearly what exactly "cellular automata" are? The only way I parse that is as: automata (i.e. machines) that are put together similarly as cells are put together in organisms. Is that what it is? Modelling things using a organism-cell model? That is, as a bunch of small machines working together?

Rachael
"Like cheese spread over too much cantelope, the people I spoke with liked their shoes." Ctrl-Alt-Del

Definition (none / 0) (#88)
by Kalani on Thu Jul 18, 2002 at 09:41:36 PM EST

You can use the dictionary.com definition if you'd like.

In essence a cellular automaton is a system made up of a number of cells (generally infinite in all directions, but finite systems are also analyzed) in which every single cell evolves according to a rule that usually accounts only for the nearest neighbors of a cell, and rule evaluation in this system sometimes happens instantaneously across cells at discrete time steps ("sometimes" because there are other models that don't have this restriction). For instance, you might take a single line of cells (so each cell only has a connection to one neighbor on each side of it) and define a rule that says that a cell can only become black if it was black on the previous step or if one of its neighbors was black. If you then start out the whole system with one black cell amongst the infinite/finite array of white cells, the effect will be that the automaton will grow homogeneously black in both directions as time goes on. Alternatively you might model the evolution of the system using a sort of time/space diagram in which the previous step of automaton evolution is shown above the current one. In this case, evolving the rule described earlier will result in a uniform black triangle.

There are many other systems that are variations on this theme (next-nearest neighbor rules, 2D schemes, 3D schemes, asynchronous time steps, network evolution, etc). This google search should be a good starting point, if you're interested.

-----
"Nothing says 'final boss' like a giant brain in a tube."
-- Udderdude on flipcode.com
[ Parent ]
Thanks (1.00 / 1) (#111)
by R343L on Fri Jul 19, 2002 at 12:09:05 PM EST

I ended up finding a review of this book that described what cellular automata are right after I posted this. :(  Sorry.
"Like cheese spread over too much cantelope, the people I spoke with liked their shoes." Ctrl-Alt-Del
[ Parent ]
William Fredkin (none / 0) (#89)
by emmet on Thu Jul 18, 2002 at 09:42:44 PM EST

I found this to be a very easy introduction to the topic- an article on one of the other folks thinking in this area, William Fredkin (who, surprizingly enough, seems to feel Wolfram doesn't properly credit his sources). http://digitalphysics.org/Publications/Wri88a/html/

[ Parent ]
I think you mean Edward Fredkin (none / 0) (#91)
by Kalani on Thu Jul 18, 2002 at 10:19:18 PM EST

Do you have a link to anything by Ed Fredkin that demonstrates the complaint you're talking about? I read the article you linked to a long time ago (and it was published in 1988, long before Wolfram's ANKoS came out). In a comment below, I pointed out a post by Ed Fredkin who, at the very least, respects Wolfram and his work.

-----
"Nothing says 'final boss' like a giant brain in a tube."
-- Udderdude on flipcode.com
[ Parent ]
Ahem. Edward... yes... (and more reviews) (none / 0) (#118)
by emmet on Fri Jul 19, 2002 at 02:17:37 PM EST

This was a list of links sent out on the phil agre newsletter:

A Collection of Reviews of Stephen Wolfram's Book "A New Kind of Science" (some interesting experiment and speculation, much ego)

http://www.math.usf.edu/~eclark/ANKOS_reviews.html

The newsletter can be found at:

http://dlis.gseis.ucla.edu/people/pagre/rre.html

Consistently superb info. Apologies for the mistake!

[ Parent ]

Yes I've read all of those (none / 0) (#124)
by Kalani on Fri Jul 19, 2002 at 07:38:23 PM EST

In fact, Edwin Clark (the person who put up the collection of ANKoS reviews you linked to) posts on the cellular automata newsgroup which I referenced in a post below. I've read all of those reviews and much of the discussion in the actual CA newsgroups by people who've been studying them for quite a while (which I consider a more reputable source than the laymen reporters who often misrepresent the nature of Wolfram's book).

Still, I don't think that either of the URLs you mention show anything by Ed Fredkin denounching Wolfram. On the other hand, the link I gave in my parent comment (which you seemingly didn't acknowledge) is actually a comment by Ed Fredkin in May of this year explicitly stating his agreement with Wolfram on the certain fundamental point of discrete time/space. You made the claim that Fredkin (an associate of Wolfram and the guy who introduced Wolfram to CA in the first place) has criticized Wolfram somehow. I'd like to see that demonstrated in some kind of commentary by him please.

I'm not saying that he hasn't said anything like that, of course. I do think, however, that it seems unlikely given the opinion expressed in the post by Fredkin that I linked to above.

-----
"Nothing says 'final boss' like a giant brain in a tube."
-- Udderdude on flipcode.com
[ Parent ]
Thanks [nt] (5.00 / 1) (#112)
by R343L on Fri Jul 19, 2002 at 12:09:25 PM EST


"Like cheese spread over too much cantelope, the people I spoke with liked their shoes." Ctrl-Alt-Del
[ Parent ]
Technical review (none / 0) (#97)
by asdf on Fri Jul 19, 2002 at 02:24:48 AM EST

Also check out the really good (but somewhat technical) review of the book here.

the end of the world (none / 0) (#99)
by jacks evil personality on Fri Jul 19, 2002 at 03:06:32 AM EST

What? Nano-particles disobeying laws of thermodynamics!!?

Sureley this means the end of the world is nigh!

// I am Jack's evil personality //

I was reading Principia Mathematica (4.00 / 3) (#101)
by ausduck on Fri Jul 19, 2002 at 05:45:53 AM EST

In his description of instantaneous velocities, Newton uses these "fluxions", which are infinitesimally small.  When one of these "fluxions" is squared, he ignores it, reasoning that it must be even smaller and can therefore be neglected.  But how can something infinitesimally small be neglected in one case and not in the other.
Isaac Newton, if you are reading this, I have two words for you:
- crack.
- pot.

I'm really curious (none / 0) (#107)
by kubalaa on Fri Jul 19, 2002 at 11:12:42 AM EST

... I've never had a teacher explain this part of integral calculus any better. Can someone here enlighten me?

[ Parent ]
Are you a programmer? (none / 0) (#114)
by Humuhumunukunukuapuaa on Fri Jul 19, 2002 at 12:44:44 PM EST

If you are, and need a better intuition of calculus, you can implement 'fluxions' as a class in C++, say, and get a really elegant way to compute derivatives of arbitrary functions.

Otherwise if you want to do it properly you need to study limits formally.
--
(&()*&^#@!!&_($&)!&$(*#$(!$&_(!$*&&!$@
[ Parent ]

Well, differential calc (3.00 / 2) (#125)
by Kalani on Fri Jul 19, 2002 at 07:57:02 PM EST

The squared fluxions referred to by the parent can simply be considered as the differential terms. The way that it generally works out when you take a derivative is that the squared differential terms, for instance actually get factored out and in the final evaluation are taken to approach 0 arbitrarily close.

For instance, take the derivative f'(x) of f(x) = x^2. The definition of a derivative for any function f(x) is:

f'(x) = lim(dx->0) (f(x + dx) - f(x)) / dx

So if you put in the f(x) = x^2 definition, for example, you have:

f'(x) = lim(dx->0) ((x + dx)^2 - x^2) / dx

Using basic algebra that expands to:

f'(x) = lim(dx->0) (x^2 + 2*x*dx + dx^2 - x^2) / dx

Which simplifies to:

f'(x) = lim(dx->0) dx(2x + dx) / dx

Then you say, "ok so I can cancel the dx on the top with the dx on the bottom as long as I allow that dx will NEVER equal 0" (ie: no division by 0 is allowed). So you're left with:

f'(x) = lim(dx->0) 2x + dx

So the whole reason for the lim(dx->0) comes in here, because it's a formal statement of the sort of offhanded clumsy statement, "well now I want dx to be 0 so that my interval of change will be effectively instantaneous ... but I promised it wouldn't be 0, so I'll let it get so close to 0 that it's essentially 0." Thus:

f'(x) = 2x

Or in general, given the chain rule (which you can prove pretty easily):

Given f(x) = x^m, f'(x) = mx^(m-1)

This analytical notation is actually a mixture of the notation used by Leibniz and a few others (including the concept of limits and so on) -- although Leibniz didn't use the primed function notation for derivatives.

Anyway, by ignoring the "fluxional quantities" Newton was essentially doing the same thing as taking a limit. Newton's definition of a derivative was actually given *geometrically* (he was primarily a geometer and considered the geometric proof superior to the analytical one). He only reluctantly introduced his so-called "dot notation" (where the degree of a 'fluxion' is given by the number of dots above the term, and the term itself indicates the varying quantity relative to which the fluxion itself is taken). So when he also gave analytical demonstrations (because the philosophy of Descartes was huge at the time, and Descartes preferred the analytical method of employing symbolic expressions rather than geometric constructs) he probably spent less time demonstrating how it all worked. Nevertheless, it's perfectly fine to say that a squared differential term goes to 0 just as a differential term itself goes to 0. If you apply the definition of a derivative above to f(x) = x^3 you'll see how it works.

If you're very interested, you can read about Newton's fluxional method, his geometric proof, and pretty much all aspects of his life in the wonderful biography Never at Rest: A Biography of Isaac Newton by Richard Westfall.

In any case, the root comment wasn't really criticizing Newton's fluxional method but rather the sort of superficial objections raised by somebody who's only got a passing knowledge of the subject matter.

-----
"Nothing says 'final boss' like a giant brain in a tube."
-- Udderdude on flipcode.com
[ Parent ]
Stephen Wolfram is not Isaac Newton [n/t] (none / 0) (#109)
by ghjm on Fri Jul 19, 2002 at 11:40:45 AM EST



[ Parent ]
Well (none / 0) (#113)
by BloodmoonACK on Fri Jul 19, 2002 at 12:38:16 PM EST

I don't know what a 'fluxion' is, but it seems to me he is just saying that when you square numbers less than one, they get even smaller, right? After all, for any positive integer x > 1, 1/x > 1/(x^2), right? So 1/inf > 1/inf^2 (not really true, I believe that inf^2 = inf? I'm not sure).

To sum up, he's saying one thing's small, but important. The square of it is thus so incredibly small it is unimportant. If any of that made any sense (I can't explain things very well.)

"It's like declaring a 'war on crime' and then claiming every (accused) thief is an 'enemy combatant'." - Hizonner
[ Parent ]

Nonstandard analysis (none / 0) (#135)
by Repton on Sat Jul 20, 2002 at 08:15:22 AM EST

There is a branch of mathematics called Nonstandard analysis, which one of the lecturers at my uni is quite keen on. It introduces the notion of "hyperreal" numbers, which are numbers you can construct from the real numbers in much the same way as you can construct the reals from the rationals (ie: a hyperreal is an equivalence class of sequences of reals).

The most interesting effect of building these hyperreals is that you get so-called "unlimited" and "infinitesimal" numbers. An infinitesimal number is smaller (in absolute value) than every real number, but greater than zero. An unlimited number is larger (in absolute value) than every real number.

The neat property of these numbers is that they allow you to prove theorems about calculus without using limits.

Anyway, I don't want to go into detail (because I'd probably get it wrong), but you can use this mathematics to provide a solid basis for the "fluxions" of Newton, and his reasoning turns out (more or less) OK.



[ Parent ]
about The Second Law of Thermodynamics. (5.00 / 1) (#128)
by nodsmasher on Fri Jul 19, 2002 at 09:05:04 PM EST

apparently it is wrong
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Most people don't realise just how funny cannibalism can actually be.
-Tatarigami
How's this for a review? (none / 0) (#134)
by jck2000 on Fri Jul 19, 2002 at 10:30:54 PM EST

1. Know something about CA, etc. but by no means an expert. 2. Bought ANKOS w/in a week of its debut for $50. 3. Read it cover to cover over the next month. 4. Was irritated by the lack of attribution. 5. Got that "What is interesting isn't new, what is new isn't interesting" kind of feeling. 6. Anybody want my copy?

Not Worthless (none / 0) (#146)
by engine16 on Tue Jul 23, 2002 at 04:21:01 AM EST

It's strange. I got a wealth of value from this book, though quite possibly not in the fashion for which it was intended. I have no sense for formulae or in truth mathematics as a whole, but what I did gleen from this deskweight is a growing awareness of my own writing style as a novelist. Bits of universal truth in the heart of discovery, more useful in my craft than the scientific search for unifying field theories and whatnot, though no less interesting from a non scientific point of view to read.

Ape Infinitum

Re: Not Worthless (none / 0) (#147)
by AzTex on Tue Jul 23, 2002 at 01:57:31 PM EST

That's interesting. Do you write in a particular non-technical field or genre?



solipsism: I'm always here. But you sometimes go away.
** AzTex **

[ Parent ]
Field/Genre (none / 0) (#148)
by engine16 on Tue Jul 23, 2002 at 08:38:27 PM EST

Fiction. Abstract adventure stories mostly.

Ape Infinitum

[ Parent ]
I scored! (none / 0) (#150)
by StephenThompson on Fri Jul 26, 2002 at 05:08:21 AM EST

I sold my copy to my roomate for $20. Sucker! I even warned him. The review is really not very good; I say this because it is far too kind to Wolfram...

You hate to quote me? (none / 0) (#151)
by trhurler on Mon Apr 28, 2003 at 03:00:42 PM EST

That's so sad for you. Quoting me gets you all the girls. You will have to die lonely now.

As for Wolfram, I suspect history will be both more and less kind to him than present day readers. More, in that there are probably a few truths in that book that weren't obvious to everyone, and less, in that you people are giving him a lot of attention he neither deserves nor will get in the history books.

Incidentally, particle physicists are coming up with scenarios that they think might locally violate the laws of thermo all the time; they just never come up with one that actually matters. Wolfram's problem is not that he fails to be a diehard empiricist, but rather that he fails to accept the reality that he might not be the second coming of Christ.

--
'God dammit, your posts make me hard.' --LilDebbie

Review of Stephen Wolfram's A New Kind of Science | 151 comments (132 topical, 19 editorial, 0 hidden)
Display: Sort:

kuro5hin.org

[XML]
All trademarks and copyrights on this page are owned by their respective companies. The Rest 2000 - Present Kuro5hin.org Inc.
See our legalese page for copyright policies. Please also read our Privacy Policy.
Kuro5hin.org is powered by Free Software, including Apache, Perl, and Linux, The Scoop Engine that runs this site is freely available, under the terms of the GPL.
Need some help? Email help@kuro5hin.org.
My heart's the long stairs.

Powered by Scoop create account | help/FAQ | mission | links | search | IRC | YOU choose the stories!