A New Kind of Science

Rating:
Summary: Disconnected
Review: Wolfram's disappointing, self-promoting book suffers from two fundamental disconnects.

First, there is a enormous gap between a basically visual inspection of pretty computer graphics that resemble something real, and any understanding of the real thing itself. Desperately short on rigor, ANKOS relies on qualitative, visually compelling imagery. So the first problem is the disconnect between imagery and reality. CA can make pictures, interesting ones, but still I learn nothing about the thing it supposedly resembles. Science? Not hardly. This is a glossy publishing job with little more substance than the Face On Mars -- another meaningless pattern we all seem to recognize as something familiar.

The second major disconnect is between Wolfram himself and, well, just about everyone else. Everything about this book points to a guy who has sequestered himself from all criticism and external review. The writing is atrocious, making it clear that no competent editor was let within one red pen's reach of the manuscript. Even more flagrantly, this book manages almost completely to avoid crediting anyone else or the existing literature on many of Wolfram's topics. This is unforgivable in a presumably scientific work. Ultimately, Wolfram's apparent disdain for other, and therefore lesser, opinions has taken this from possibly being a science book (for which peer review, criticism, and reverification would offer it credibility) and turned it into little more than an opinion piece.

Still, I occasionally hear people say that this is an authentically new book, that Wolfram is a courageous genius turning the Scientific Establishment upside down, and that's why he's chosen to go it so obviously alone. I would suggest an alternate explanation.

Surprisingly often, people become convinced that they and they alone are the True Genius of The Age, typically seeing profound universal interconnections others have failed to recognize. They begin to assume eccentric habits, withdrawing from others while suffering simultaneously from a delusion of their own special importance. If such people are poor, they are eventually diagnosed and subsequently medicated. If they are rich and have a formidable scientific resume, they write a book like this.

Rating:
Summary: Dimension of the Debate
Review: Read this book if you are a psychologist or sociologist and you want to see first hand how a man's ego can get out of control if he isolates himself from society for a decade.

What is presented is not science, but a form of math.
The concepts of cellular automata are not new.
As a result, the book's title should be "An Old Kind of Math".

The algorithms shown are correct and there are some new and very interesting observations. The debate over a book like 'The Origin of the Species' centered on the truth of the content. The debate over this book centers not on the truth of the algorithms presented but over the ego-distorted orientation of the author.

Rating:
Summary: Five Stars for Courage and Content
Review: Though difficult to imagine even decades after the fact, and even more unfathomable after centuries, it is nonetheless true that individuals who made great contributions to humankind were initially ridiculed, impeded, and persecuted.

Some quotes which apply:

"Ridicule is a public confession of fear." - Vanna Bonta

"Great spirits have always encountered violent opposition from mediocre minds." - Albert Einstein

"Calumny kills more men than swords." - Leonardo Da Vinci

"Eppur si muove." Galileo Galilei

It's always exciting to find a work that quickens the pulse.
Reading these reviews is sad evidence that lynching individuals for works of orginality is still an unfortunate human reaction among the mob-entwined. In Leonardo da Vinci's personal journal, he laments being snubbed and scorned by scholars of the day. He explains that while he cannot quote as they do, the pages of texts, his teachers are the teachers of their teachers: Nature and observation. If only for his dedication and courage, Wolfram merits at the very least respect for his effort. The text also contains some interesting axiomatic principles that redeem discovery from the mire of cogitative thinking. Rather than approach the problem through the equation, he first envisions. Simplcity, of which Wolfram is a proponent, is more illusively grand than the intellect can possibly fathom.
Thanks to the author for his independent sacrifice and dedication. Sympathy to the more ambitious who wish, for reasons of self importance, they had done the same. They try to trash the book, some before having read it.

Rating:
Summary: A New Kind of Science
Review: Like many readers I was disappointed by this book (although I have not finished it I have read large slabs and skimmed the rest). What I was looking for was the ability to take something complicated and then use the theory to deduce its "generators." My initial read of the book suggests that the closest one can get to this is a comment along the lines that you could test a small (or hopefully small) number of "classifed" structures to see if they produced the outcome. However, there is no systematic way of how to do this as far as I can tell. This is not to say that someone might not make some advances in that area.

While there is an "ooh, aah" factor about some of the graphs, unlike some other reviewers I think they could stimulate people in other disciplines to think about their problems in new ways eg nanotechnology, biology etc. That is not a bad thing. The curious thing is that some of the graphs are so evocative of actual structures that one sees in electron microscope scans that one feels that there may be some deep linkage which could have predictive value. It's the elucidation of that potential that would be truly fundamental ie something that shows in a predictive sense how elements with different atomic weights etc will be structured and how combinations would be structured.

Obviously Steve Wolfram has got up a few noses for some reason judging by some of the rather stentorian criticisms I have read. I will plough on with the book on the basis that it may provide some insights that can be developed further. While it may not be Principia Mathematica in terms of its completeness, novelty and relevance, if a handful of the right people get some insights from it and develop them it will have proved a useful contribution notwithstanding the reservations.

Rating:
Summary: Exhaustive study of 2^(2^(3^1))=256 mappings.
Review: This book gives an exhaustive study of 2^(2^(3^1))=256 mappings that generates one-dimensional cellular automata. Do the discoveries or methodologies in this book generalize to 2 or 3 dimensional cellular automata? Is the author suggesting doing the same kind of exhaustive study of 2^(2^(3^2))=1.3408x10^154 mappings for 2-dimensional or 2^(2^(3^3))=2^134,217,728 mappings for 3-dimensional cellular automata with a regular grid of neighbors? (m^n means m to the n-th power, if you are used to a different notion.)

Rating:
Summary: It's all in the patterns
Review: An enlightening and fresh look at the world of science--beautifully written. Highly Recommended.
Also highly Recommended in this genre: "Decoding Darkness" by
R.Tanzi and A.Parsons

Rating:
Summary: A "naked emperor" kind of science
Review: This book would be completely ignored were it not for Wolfram's "child prodigy" pedigree and his deep pockets (as if either mattered --- how many advances are made by prodigies or by the wealthy?). Wolfram didn't invent cellular automata and the notion that simple rules can yield pretty patterns that sometimes look like natural phenomena is old news. Even the illustrations are boring --- despite the hype about how much care went into the printing of this book, much better illustrations of the same type can be found in Mandelbrot's fractal manifesto published decades ago (and in color!).
Wolfram's inability to credit other authors is inexcuseable, and it's insulting to compare this self-centered screed to works like "Origin of Species" or "Principia Mathematica" --- Darwin copiously credited others for the work that preceded him and Newton acknowledged that he saw further because he stood on the "shoulders of giants". Apparently, Wolfram doesn't need to stand on any giants, but his vision is quite limited as a result.
Wolfram will dismiss criticisms as narrow-minded, but no matter. Great works will stand the test of time, and garbage will sink to the bottom. Given the weight of this tome, it will sink very quickly and very deeply.

Rating:
Summary: REAL GENIUS
Review: Mr Wolfram is really a genius . Only a person with this gift is capable of selling thousands, if not millions,coppies of a worthless book

Rating:
Summary: Good book but misleading: a cellular automata encyclopedia
Review: It's ironic that this came out at about the same time as the late Stephen Jay Gould's opus on evolutionary theory, because both of these very brilliant guys seem in a sense to be guilty of the same crime, posing in an exaggerated way as scientific revolutionaries. Gould is widely considered to have exaggerated the orginality and usefulness of things like "spandrels" and "punctuated equillibrium," while Wolfram seems to want to take credit for applying computational complexity science widely to nature. Both men do make good points, and both seem to succomb to their own sense of exaggerated importance at times.

Ok, so I fell victim to mass hysteria and read this. Although I have only a passing interest in cellular automata, I felt I had to read this because of all the hype around it recently. Well, I'm not an expert in CA, but it seems like a nice encyclopedic text for those who are. I found most of it boring. It reminds me of those artificial intelligence texts that sound like they are going to reveal how machines can be made to think, only to find an encyclopedia of subroutines.

I think people who write books like this seem to confuse two different kinds of audiences: people who want to hear a lucid explanation of how something works, and people who want to read code. This one is better oriented to the code readers.

From my perspective, this is not particularly interesting or new as philosophy of science, or "a new kind of science" as the title seems to imply. It may seem a little unfair to blame Wolfram for the excessive expectations of readers, but then again he does seem to encourage the view that he is presenting something completely revolutionary.

The view here is a reasonable modern extension of the Turing-Church theorem, but applied to the natural world. Wolfram demonstrates that cellular automata can often also be "universal computers," so that simple algorithms applied over and over and over can not only yield patterns of great complexity, but most importantly, like Dan Dennett's similar view of Darwinian selection as a universal acid etching its way through natural design spaces, the process can solve real problems under the right conditions.

Natural processes can be at least largely described and explained by how they build themselves up through repetitive rules. I thought Holland and Kauffman and the others had been demonstrating this for years, and it seems to me to make more sense to consider Wolfram's work another step in that program rather than something qualitatively different. In contrast, Stuart Kauffman's "Origins of Order" gave a similarly encyclopedic view of his work, but while crediting the sources of the ideas and also providing a realistic perspective of what was unique to his work and where he was building on other peoples' work, and giving a sense of contrast.

Wolfram's computations are indeed a new concept of mathematical science compared to Descartes, Newton, Laplace, and Bernoulli. These patterns build themselves up rather than winding down to entropy. However it isn't new with Wolfram, it was new with Turing, Church, von Neumann, Ashby, and Weiner, among others, and developed into computational research programs by the entire Santa Fe Institute.

As an introduction to the computational view of nature, this book is very weak, since it doesn't give the intellectual foundation, just the author's view.

So long as you know what to expect, this is a useful text of cellular automata and a good addition to the "new science," just don't buy it soley because of the hype around how utterly revolutionary it is supposed to be, or because of how Wolfram's personal brilliance has been emphasized, or I think you are likely to be disappointed with what this book really contains.

Books I prefer as introductions to "the new kid of science" of computation and pattern formation in nature include Phillip Ball's "The Self-Made Tapestry," Holland's "Hidden Order," and Waldrop's "Complexity: The Emerging Science."

Rating:
Summary: Yuri Kuzyk, please give more detailed references
Review: This book is worth a read for the simple fact that it has already generated a lot of debate.

I'm curious to follow the references mentioned in Yuri Kuzyk's review, however I think some of them might be journal articles, and I'm having a hard time tracking them down.