A New Kind of Science

Rating:
Summary: DOES THE META-NARRATIVE OF THINGS REALLY EXIST?
Review: There have been plenty of comments about the thick book (about 1,280 pages), but certainly not because the thickness. Plenty of responses to the book because it was written by the Stephen Wolfram, probably the most excited scientist in the world campaigning for cellular automata all at once making fortune from it. One of Wolfram's great contributions to science is developing practical symbolic computer algebra and wide spreading for the first time, the software MATHEMATICA. The other thing for so many people talk about the book is because somehow, it brings some controversies in the society of advanced science in the world of complexity science.

Cellular Automata (CA) is arguably very nice tool to use for analyzing many things, but I don't think that CA itself the new model for explaining the world. CA has been formulated since the original works of John Von Neumann, and there has been so many development and improvement of the CA since the vast technology of microcomputer recently. Furthermore, I don't think the label of "A New Kind of Science" is not very nice for the ideas representing by the book, as if the book is just the same level with historically the revolutionary works by Emmanuel Kant, Isaac Newton, or even Galileo.

The most interesting part of the book is that the book is full of beautiful figures of the cellular automata. As we know, Cellular Automata (CA) is the spatially and discrete computational device that initially begins with extremely simple rules but creates interesting images which show some even more interesting behaviors. These one dimensional cellular automata images are a bunch of black and white pixels that are built from the top down, where each scan-line is based on the color of the pixels on the scan-line above it.

An intriguing claim from Wolfram in the book inviting so many criticism is that when he saw the complex patterns of his CA, he claims that all of the pattern is as universal as how the nature behave on pigmentation. This invited biologist's criticism, as there is no further researches on it. It is a computational simulation of the nature and the simulations shall use some reductionism methodologically while Wolfram talk in epistemologically speaking.

However, the book is very valuable on giving any explanation on the possibility of using CA explaining our world. As we know Wolfram did many researches on how to explain very hard things such as the thermodynamics law in CA-style, a very superbly incredible thing to do. The book should be read but not only this book, I concur, as the book seems to be too egocentric without the bibliographical references acknowledge. However, in the web-site of the book (http://www.wolframscience.com) some noted for suggested reading is submitted.

Last but not least, it has to be acknowledged that the book will give a fresh air to the traditional scientists with traditional methods and paradigms - a book challenging the natural scientists and opening the eyes of the social scientists on the contemporary usage and articulation of signs. Eventually, I suggest you to read it and simultaneously surf in the web-site prepared for it as you will find many interesting and inspiring ideas on how we will shape the world of civilization.

Hokky Situngkir
Bandung Fe Institute

Rating:
Summary: Stunning and full of wonder
Review: Many of the reviews of this book have been quite savage, but the savageness, it seems to me, is inversely related to the fraction of the book that the reviewer had actually read. I, like many others, was aware of the marketing hype before its publication, and waited for its production with a feeling something like that of a child anticipating a Christmas, an incomprehensible number of weeks off.

I actually read a good part of this book, and found it to be something quite awesome. Moreover, I was not at all put off by its idiosyncratic style, and I much appreciated its typography, layout and the richness in its illustrations. I was awed at the scope of the work, much provoked and entertained by it. A faithful user of Mathematica (I used it to derive my first simulation of the trebuchet, described at www.algobeautytreb.com), I am awed by the number of programs that Wolfram lists.

Perhaps the most impressive aspect of this work is the breadth of topics that are covered. Computational hydrodynamics and thermodynamics, free will and language, cryptography and logic, quantum mechanics and Feynmen diagrams, crystallization, coarsening foams and sandpiles, evolution and embryology, traffic and financial markets. Is there any other work having a single author that encompasses such a collection?

The notes section, about 300 pages long, is wonderful to read all by itself. Each topic has its bit of history related, and there are many descriptions of things discovered and things yet to be discovered, told in a most entertaining and readable style. I could even recommend, perhaps, that some readers should read this section first.

My reaction toward the "bottom line" hypothesis, that cellular automata provide a real breakthrough in understanding the world, is a little agnostic, perhaps. Interesting in many ways, and the hydrodynamic model is clearly a fruitful one--I've seen some quite stunning movies of certain phenomena (such as two liquids of different density that are mixing). I really can't evaluate some of the claims Wolfram makes, and others seem to be a bit overblown. I guess my feeling is something like the child that woke up on Christmas morning hoping for that bicycle, but finding a chemistry set instead. While I didn't get exactly what I'd hoped for, the chemistry set is intriguing--something may come of it yet!

Rating:
Summary: Mr. Wolfram has plagiarized the work of Richard Dawkins
Review: The book has good ideas! Unfortunately Mr. Wolfram
has not realized that a number of conventional scientists have proirity over his published work and ideas. Mr. Wolfram has not referenced any of them and this by definition is PLAGIARISM ! Mr. Wolfram should only get credit for popularizing the ideas in his book. For a start read "The blind Watchmaker" by Richard Dawkins. Richard Dawkins has done most of the experiments mentioned in MR. Wolframs book. Mr Dawkins has elucidated almost all of the ideas which MR. Wolfram thinks are his own original invention.

Rating:
Summary: A breakthrough, a bomb for the shallow minds
Review: Nature is very simple at its most basic level. That is what this big book written by outstanding genius says.
The universe is one fractal like interaction, which generates at the scales of observation what we see as space, time and everything else. This is what another great scientist Eugene Savov claims. He confirms the findings of Stephen Wolfram by revealing the extreme simplicity of the structure of reality - its 3D-spiral code.

To put it simply those who neglect the underlying simplicity of nature are blind for the principle of parsimony and for the history of science in which A New Kind of Science and its author will remain.


Rating:
Summary: This guy is absolute egomaniac.
Review: Who edited the book? This book has 2-3 "I's" and "me's" in a sentence. It was disgusting. I discovered this I discovered that, without me scientists couldn't do this or that. Not even a drop of simple modesty in that book. Don't waste your time and money on this book (by the way the way it's printed is absolutely wasteful the font is big the line breaks are huge the book has 1200-1300 pages that you can easily fit in 400page book) Anyway buy a good math or physics book don't waste your money on this one.


Rating:
Summary: A new kind of bubble: elementary computer science as a TOE
Review: I am among those who admire Wolfram for having been a prodigy, and especially for having created Mathematica. No doubt, he is a very special person - a talented businessman and an excellent computer scientist, too. He also started as a promising, very young physicist. These are the reasons why I gave the book an extra star that it would not deserve otherwise.

Many people have criticized Wolfram for his pretending that he is the inventor of all these rather standard ideas and facts in computer science. I think that these critics are right. But enough has been written about this aspect of his work. Moreover, computer science is not my field anymore.

I want to say a couple of words about Wolfram's ambitions to apply these ideas to natural science - a new kind of physics, so to say. I am surprised how the reasoning of such an exceptional mind can become so superficial, narrow-minded, cheap, and isolated from reality and the ideas of others.

Wolfram seems to be very impressed by that simple algorithm that produces an irregular, "unpredictable" pattern. The physicists like me just don't understand why. How can such a smart person be impressed with something so common? Most systems in physics (and science) that can be written down cannot be solved exactly - we usually say that they are not integrable. Even the system of three massive bodies (planets) that obey the simple laws of Newton lead to motion that cannot be expressed in terms of "ordinary" functions.

It is not shocking to find a system whose behavior looks irregular and unpredictable, even though - I realize - most of those 256 simplest cellular automata are integrable. On the contrary: the physicists are usually impressed if someone shows that a set of equations *can* be solved. They are happy if someone shows that a seemingly irregular pattern exhibits some deeper rules.

The cellular automata are nice toy models in computer science, and they are similar to discretized models in classical physics (namely classical local field theory). But that's it. Classical field theory is, roughly speaking, an achievement of physics of the 18th century. Thousands of new and amazing ideas have been found in physics, especially in the 20th century.

Wolfram seems to see one idea only - the idea of the cellular automata. The output of a simple computer program looks like a piece of tiger's skin - and it is apparently enough for Wolfram to think that his program, or something very similar, probably contains all of science including biology. Well, it is obvious that these simple models can never agree with the pillars of modern physics, such as Einstein's relativity (1905) and especially the principles of quantum mechanics (1926). They are what they seem to be: simple material for students of computer science. Physics - and even Mathematica - contains many more organizing ideas and structure that is necessary for them to work.

It seems to me that several decades ago, Wolfram had to know all these fields of physics very well. What happened afterwards? Why does Wolfram suddenly say, much like a generic crackpot, that everything is encoded in one, rather naive idea? Does he really believe that the content of ANKOS is so important, or did he just want to earn some more money? I am not sure which answer is more worrisome.

Rating:
Summary: Upon a Second Read, More Favorable Opinion
Review: When I initially reviewed the book, I gave it 5 stars for what I considered groundbreaking work.

But then I began to consider that perhaps there were some missteps, so I downgraded to a 4-star rating.

The 4-star rating was granted based on these pros and cons:

PROS:
1) The author explains his central claims about intrinsic generation of complexity more clearly and more correctly than other authors.
2) The breadth of the content, particularly in the notes section of the book, allows one to understand the history leading up to A New Kind of Science as well as the author's need to address real and/or perceived problems with the current general scientific framework.
3) The Mathematica programs and their output used to generate the diagrams are mostly original and challenging, but understandable for a competent computer programmer. Most of the diagrams, as well as the non-technical explanations of the diagrams, are moderately difficult.

CONS
1) The appendix lacks formal citations.
2) Some phrases are overused throughout the book.

After I read the book as well as the positive and negative critiques of the author's work, I went on something of a scientific literature exploration. I read the works of authors referenced in the book, as well as the supposedly copied-from works referenced primarily by negative critiques. I felt confident after roughly a year of extensive reading that I understood most of the relevant literature as well as the epistemological issues surrounding the book. My deeper exploration confirmed my original assertion that the author's theoretical claims are indeed original, contrary to many negative critiques from media, "regular people", and some scientists. As an added bonus, I had a lot more to think about.

So I started to read the book for the second time to further check whether I missed something either on my fact-checking or even in just understanding the book in general.

I am nearly finished, and I must say I have found the book much more rewarding this time around.

Today, I reinstate the 5-star rating.

Why?

I now understand something I didn't really grasp the significance of before: the author has argued much better than others why matching constraints is not a good way of doing science for many difficult problems. The "old kind of science" thinkers could learn from the author that program sampling is a valid way of doing things. That is not to say that scientists were not sampling programs before the book. The author just "took it to the next level" through providing a new way of reasoning about these programs.

My early criticisms of the lack of citations and the overuse of phrases before were unjustified. On the citations, it would be a waste of paper to point to all of the documents rightly associated with this book--there are simply too many. Nonetheless, I think it would be nice to see an online repository of full citations, if such a thing is possible. On the overuse of phrases, I take it back because the way the author crafts his language is intentional for keeping his framework valid. He states his claims such that they are difficult to refute, and this is where the perceived redundancy comes in-and, I would argue, some people's disillusionment with the author. People may not like the author's style of writing, but at least he does not lie when he makes theoretical claims.

I would say buy the book. Take note that you will get much more out of the book if you understand general analytical methods from mathematics, computer science, and physics. That means you should really have knowledge in multivariable calculus, sets and their operations, thermodynamics, non-Euclidean geometries and relativity, algorithms, functional abstraction, recursion and so on. A high-level understanding of quantum mechanics will not hurt, either. As far as issues in biology are concerned, an understanding of genetics and cell biology will be helpful. In addition to the a strong grasp of the concept of abstraction, knowing how to use the tools from multivariable economic theory, cognition theory, and sociology will be particularly useful when encountering the author's formulation of the Principle of Computational Equivalence.