A New Kind of Science

Rating:
Summary: A new kind of egotism, but ...
Review: ...but a fun (though inconsequential) book nonetheless.

The book's theme is that highly complex processes can emerge from the interaction of simple parts according to simple rules.
This is convincingly illustrated using many examples of cellular automata. In fact, I give the book three stars because the first chapter has a nice general introduction to the theory of cellular automata.

I give it *only* three because the ideas aren't particularly new, and the chapters on specific topics (e.g., visual perception) are not only not new, but out of touch with present-day science. For instance, The view that complexity can emerge from simple interactions has been around since von Neumann invented cellular automata in the 1950s, so this kind of science is not particularly new. Certainly, with the advent of computers that have the processing speed to carry out these simulations automata theory has progressed greatly since then, but this is not much more than putting a nice dressing on an old display.

Another problem with the book is that its central claim is most likely *wrong* (i.e., the claim that the universe is a cellular automaton). For one, certain complex objects, such as the human brain, do *not* use simple rules of interaction: complex network properties of nervous systems emerge from the *complex* interaction of *complex parts* (i.e., neurons, each of which exhibits extremely complex biophysical properties).

Anyway, despite the ridiculous amount of hype Wolfram has generated, I can recommend the book for anyone interested in basic automata theory. If you are a practicing scientist, don't burn your textbooks yet. Wolfram has a lot to learn from them.

Rating:
Summary: Disappointing
Review: If the author refrained from using the word "I", he could have shortened this work by nearly a hundred pages. This book screams "HIRE AN EDITOR".

I was interested in Wolfram's underlying ideas. But could barely discern what those ideas were between the self aggrandizement and atrocious writing. It becomes obvious very rapidly why he didn't propose these ideas to the scientific community at large; he would have been reviewed.

Wolfram's lead in claims incredible depth and application, but the book doesn't deliver. I was not only disappointed but angered by the author's tone and style. If he had stuck with the facts, the book would have been interesting, readable, and probably about 300 pages. As it is, it is hardly worth wading through it.

Rating:
Summary: This book is in need of an editor.
Review: This book is in need of an editor. The ideas are substantial and significant. However, the writing style is annoying.

Rating:
Summary: Wolfram's Science is Not as New as He Thinks.
Review: There is no question that Stephen Wolfram has made a contribution to scientific knowledge with his book. However, I would not characterize it as a new kind of science. The theme that runs through his book is that complexity in nature can come from simple programs that run through a large number of iterations. Recently, I gathered a broad range of information for my book "The Meaning of Life If Life is a Journey You Need Good Directions" that looks at what truths we can learn from science about reality. Based on this study, I can say that Wolfram is far from breaking new ground. Daniel C. Dennett introduced the same basic ideas in his book "Darwins Dangerous Idea" in 1995. In fact, Dennett uses as his reference Charles Darwin's ideas on evolution published more than 130 years ago.

According to Dennett it is possible to have "design without intelligence". In other words,a simple algorithm, another word for a program, can produce complex results without the system that is running the program having any prior intention of the results that are eventually produced. In his book Wolfram wonders how science could have taken so long to see things that he now sees as being very intuitive. Dennett answers this question as well. According to Dennett it is human nature to believe that nature works the same way that humans think.

Humans are the first species to have the capacity to anticipate the results of its actions. Humans can make causal connections between actions and their outcomes. According to the anthropromorphic principle humans assume that any higher power (God) that created and controls nature works under the same principles. However according to Darwin and Dennett nature is simply adapting to changes in the environment as they occur.

Wolfram states science has assumed that laws of nature were complex because we experience complexity in nature. In one sense believing that their is complexity in nature gives us a sense of comfort because we assume that something must be in control of all the complexity. If we concede that everything is subject to random events and simple programs, we also have to concede that the world is a chaotic place.

Wolfram further states that he has spent the better part of his life working on this book and we have to thank him for his efforts, because he has run the computer simulations and analyzed the data to provide conclusive proof that it is fact possible to have "design without intelligence". Furthermore, he has developed a set of principles that can be used by scientists, which could be characterized as a new scientific method. For example, when looking at a system try to reduce it to the simplest level possible. Without getting to the most simple level we may be incorporating assumptions in our study, which we may assume are intuitive, but once the imulations are run we see that are assumptions are not as intuitive as we thought.

Wolfram's book is a very long book, 846 pages without the notes, and then another 350 pages of notes. All of this information is provided to document how science, nature, etc. conform to his basic concepts. However, I am surprised why Wolfram thinks that these ideas should be such a revelation. It is generally accepted by science that the universe started about 14 billion years ago simply as energy. All subsequent matter and complexity in nature flowed from this simple beginning. Thus, we have known for some time that all complexity has flowed from this original simplicity. Any laws that we develop today should only be a natural description of this evolutionary process. Anything else would be a ontradiction in logic, at least to the extent that our limit capacity to understand nature would indicate.

Rating:
Summary: In The Beginning...
Review: Irony: The author claims that infinite complexity may arise from ultimate simplicity, yet he needs more than 1,000 pages to explain his theory.

The idea that a few lines of code created the whole universe is about as likely as the theory that we evolved from some primordial swamp. It's out of touch with the thinking mind and observation of the reality we see around us.

Whether shrouded in a thousand-plus pages or millions/billions of years, sheer volume is a great cover for nonsensical, whimsical thought.

Our world is full of examples that things complex DO NOT arise from things simple unless there is a designer and effort at work behind the scenes.

For example:

* Computer hardware by itself is useless; it needs instructions from outside the system - it can't code itself and never will, no matter how many eons you let it sit there idling and wasting energy.
* Left to their own devices, vintage autos rust; they don't magically transform into today's luxury sedans complete with entertainment centers and GPS mapping systems.
* Species are going extinct; they aren't coming into being.
* An intricate Swiss watch will never assemble itself by simply putting the pieces into a bag and shaking it - even if you shake it for a few billion years.

By considering the operation of natural law as we see it today and taking into account mathematical probabilities, it's almost impossible to arrive at any other conclusion for origins than the timeless and simple few words found in the biblical account of Genesis 1:1.

Rating:
Summary: Incredible - A simple idea that is completely unique
Review: Such a simple idea.... It is incredible no one ever thought of it before.
Rather than try to explain the complexities of the universe (like Einstein or Newton) why not explain why the univese appears to be so complex.

This book is brilliant. Makes you think long after your done reading.

Rating:
Summary: The Core Issue
Review: The book fails to explain how the automata should be designed so that they evolve as do natural entities. Thus cellular agents have no connection to useful work. Wolfram does not even come close to this issue, the only central question in the automata "field".

He only plays, as some rich scientist who has run out of ideas might play.

One serious 10 page paper on how to model such agents for useful purposes would be worth 10 such books.

Rating:
Summary: Would have been revolutionary if published in the 1800's
Review: The controversy over Wolfram's A New Kind of Science is an invention of journalists (or, as one reviewer aptly put it, "dilletantes") and others mostly outside of the scientific community who insist that no-one could possibly have the background to read and absorb Wolfram's opus well enough to critique it; surely any criticism must be motivated by petty jealousy passing for understanding. It's unfortunate that this tome was written by someone so wealthy; the wealthy have authority--even scientific authority, it would appear.

Here's the effect that Wolfram's work is having on science: professional scientists, mathematicians and philosophers are now expected to weigh in on Wolfram at cocktail parties. Uninformed snap judgements about who is open-minded and who is a petty, mean-spirited, calculus-mongering dinosaur will be determined on the basis of one's attitude towards Wolfram; body language will be scrutinized for signs of insincerity, or worse. Any criticism, however mild and however justified, will be characterized as stentorian; platitudinous cautions to ignore Wolfram's habitual self-aggrandizment will be uttered to those who presumably haven't understood the master's message, by those who haven't understood it either. Everyman is now rich with Wolfram's good-as-gold, prerequisite-free science; bums in the street will implore passer's by, "buddy, can you par-a-digm?" Wolfram's text has been a bonanza for the anti-science put-down artist (the individual whose chief contribution to the world is the put-down). The foregoing more or less characterizes the effect Wolfram's book is having on the scientific enterprise, aside from encouraging computer experimentation, here and there.

At least the designers of Maple (a competitor of Mathematica) have had the good sense not to respond to A New Kind of Science with their own textbook on cellular automata.

Rating:
Summary: Inside Joke?
Review: The last time anybody tried to base physics strictly on the phenomenological ontology of measurement by rods and clocks he came up with special and general relativity.
But this was the last time anyone saw such a program through to a satisfying conclusion, and that was almost a hundred years ago.
Since then, as most physicists will admit privately, it has become an embarrassingly dirty little secret that physics has disappointingly metamorphosed from a principled into a merely descriptive science, sometimes, as in QED, QCD, the electroweak, M and other gauge based theories, brilliantly descriptive, but at best no more than a breathtakingly accurate mathematical cataloging of observed phenomena, with the most widely heralded verifiable predictions coming not so much from theory but rather more from gaps in its known Mendelevian-like catalogue entries.
In this spirit Stephen Wolfram's gargantuan collection of computer generated Rorschach symbols advertised as A New Kind of Science might be regarded not as the apostasy decried by his detractors nor quite the breakthrough applauded by his admirers, but instead something like an heroic reductio ad absurdum to much of what has been going on in the physics journals for at least the last fifty years --in the end, given the man's acknowledged brilliance and demonstrably infinite capacity for painstaking effort, perhaps even intentionally the inside joke to trump all other inside jokes.
Shouldn't we expect more from the genius of a Stephen Wolfram?

Rating:
Summary: Wolfram's "New" Science Does Work!
Review: Stanley Grossman overlooks something. The very fact that a program like Mathematica exists, shows that you can express mathematics in the form of computer programs. Even ONE program is able to contain about 90% of all of the discovered techniques of mathematics (not the insight, but the techniques. Insight requires a human mind). From this you can conclude, that you can capture the calculational aspects of mathematics in computer programs.

Is the converse also possible? Is it possible to capture every possible form of output of computer programs in the form of mathematics? The book makes a very strong case that this is NOT so. This means, that computer programs, seen as a tool for describing the world, and thus as a new paradigm, is MORE POWERFUL than mathematics.

Stanley Grossman apparently has with his words 'at its most profound level' and 'In the final analysis' only MATHEMATICS in mind. He apparently believes that mathematics is the only candidate for 'the most profound' and the ONLY thing capable of making 'final analyses'. Apparently he does not want to be dependent on computers to see 'profound levels' and make 'final analyses'. He finds dependence on pen and paper to be going far enough.

The point of Stephen Wolfram is, that you can use COMPUTER PROGRAMS as a successor of mathematics. Mathematics being then a special case. His point is, that mathematics might be a subset of a more inclusive science. It is only unfortunate that Stephen Wolfram calls it a KIND of science, and does not see it as a full-fledged science.

I suggest that we acknowledge this to be a new SCIENCE, Period. I suggest the following name for this science: COMPUTICS.

COMPUTICS; the science that uses COMPUTER PROGRAMS as a new way to formulate testable hypotheses.

COMPUTICS in this sense of the word is then comparable with MATHEMATICS: the science that uses ABSTRACTIONS as a way to formulate testable hypothesis.

PHYSICS: the science that describes existence in mathematical form. You can have MATHEMATICAL PHYSICS, which is, essentially, all of physics.

In the same way you can have COMPUTICAL PHYSICS: the science that describes existence in the form of computer programs; you can have COMPUTICAL ECONOMICS, COMPUTICAL BIOLOGY... etc. By giving the new science the name COMPUTICS we can become aware that when we make computer models we are, essentially, doing something new. It is a new paradigm, that is a SUCCESSOR of mathematics, just like mathematics is the successor of language.

You CAN make predictions with Computics that ARE repeatable. How? By just RUNNING THE PROGRAM that embodies your hypothesis. Every time you apply Rule 30, or Rule 110, with exactly the same initial conditions you get EXACTLY THE SAME OUTPUT, although these structures cannot be described in the form of mathematical patterns.

(That is why it is not the same as Chaos Theory.) Also, because computer programs consist of COMPUTER CODE you can say, that you ARE making definite statements about reality. They are just not in the form of ordinary language, logic or mathematics. They are in the new form of computer programs. Of course, your mind alone is no longer enough. As a computician you do NEED a computer to make the descriptions and predictions, for without it you cannot write and run the programs. But, then again, a mathematician also needs the 'crutches' of pen and paper to be able to make his descriptions and to calculate predictions. His mind alone does not suffice. So there is no essential difference here also.

Also the 'inverse problem' is not only 'far more intractable' and 'practically impossible' as he asserts. It is even TOTALLY impossible to solve, even in principle. All progress in science began with HYPOTHESIS that are testable and TESTED. You are never investigating whether some statement is true, but whether it is not false, which is not the same thing. That is the major discovery of Popper.

Ever since Kant's 'Critique of Pure Reason' we know, that although our understanding BEGINS WITH existence, it does not ARISE OUT OF existence. Ever since Popper we know, that truth is not found by deriving it from the data of the senses by some form of induction. No, we make progress in the form of hypotheses and refutations. We make hypotheses, and then TEST whether they are not CONTRADICTED by the data of our senses. (Or our experiments.) If they do, we discard the hypotheses. If they do not, we uphold our hypotheses, and acknowledge at least that there IS a domain of validity.

COMPUTICS is definitely a new science: it gives us a new way to formulate testable hypotheses. We have now four ways to formulate testable hypotheses. We can make them in the rather inaccurate form of 'ordinary language', in the more accurate but still imprecise form of logical symbolism, in the accurate AND precise, but also simplifying form of abstractions (=data reductions) of mathematics. And now we can realize that computer models gives us a way to formulate testable hypotheses in a new form. We can give them the new form of accurate, precise AND complex form of output of computer programs. They have not the mathematical constraint of simplification. ALL interactions can now be taken into account, without simplifying assumptions. THAT is the EXTRA what makes Computics to be the next step in human (r)evolution.

Since I realize this, I have given this book an extra star compared to my earlier review. I find that the book 'grows on me' in importance.