Stephen Wolfram

[schu]

Anybody got the skinny on this guy? He has many of the marks of a fraud, but I'm not able to evaluate his work. He claims to have changed science as we know it. Any truth to that? Science News has an article this week that is only a bit skeptical.

I can tell that Joseph W. Newman is a fraud, but this guy is way over my head.

[echidna]

From what I recall he's not so much a fraud as just over-rated, especially of himself. He's got some new ideas, but rather than being "A New Kind Of Science" they're really not that revolutionary.

[Godless Wonder]

I think he's the main guy behind most of the code in Mathematica (http://www.wolfram.com/products/mathematica/index.html) so he knows math, it would seem. His "new kind of science" is basically all about finite automata. There's a giant book he wrote about it "a new kind of science" you can find in barnes and noble or on amazon. I think people were expecting something more from this book prior to its release (or something less "off the deep end" for lack of a better term) based on his reputation from Mathematica. He also has some books about Mathematica too. I haven't read any of them.

[Wounded King]

Wolfram certainly isn't a fraud, a rampant egomaniac perhaps but not a fraud.

[godlessmath]

He's a quack, so long as New Kind of Science is concerned. I went to a talk of his here in school, and his research struck me as rather trivial. There was also an article about his work in the American Mathematical Society which was very unflattering about his work.

He is also supposedly British, but gave me the impression that his accent was fake.

[Wounded King]

He has worked in America for the last 20 or so years, his accent could easily have deviated in that time.

I'd agree that the work doesn't seem as earth shattering as he and his supporters claim, unless of course he could find some way to demonstrate that his cellular automata are anything other than versatile simulations.

[Jesse]

Here's a useful page:
"A Collection of Reviews of ANKOS and Links to Related Work"
http://www.math.usf.edu/~eclark/ANKOS_reviews.html

The thing that makes Wolfram different from a true crackpot is that he does differentiate between speculations on the future significance of cellular automata and results which he (or one of the people working for him) have actually managed to demonstrate. The problem is that his speculations are very...speculative...and that when you take them away, there's nothing really that novel or earth-shattering about what he shows. And he's also very self-aggrandizing, and doesn't give enough credit to others who came up with significant results in the field.

[schu]

Thanks to everyone who posted for the information provided. That is just the kind of input I was looking for. Seems unlikely that he is the New Newton.

[Godless Dave]

Newton was a pompous ass as well, but he had the redeeming feature of having brilliant new ideas.

[Undercurrent]

Originally posted by Jesse
And he's also very self-aggrandizing, and doesn't give enough credit to others who came up with significant results in the field.

This is what I've heard most about the book -- He takes a lot of credit for stuff that other people did. All he actually accomplished himself was a few technical points and a lot of "These cellular automata is counterintuitive. Some aspects of the real world are counterintuitive. Therefore the universe must be underpinned by cellular automata."

[wade-w]

There was a thread on this a while back:

A New Kind of Science (http://www.iidb.org/vbb/showthread.php?s=&threadid=37916)

I think CardinalMan's input is especially relevant, since this is his area of research.

[Starboy]

I guess only time will tell if Wolfram is a quack, a fraud or a genius. I have read parts of his book. Whatever he is, he is very full of himself. What I saw was a great deal of mathematics but very little science. Just as in string theory, when the day comes that his work provokes a great deal of experimental work that supports his claims then I will think he is onto something. Until then I will think he is another Pythagorean that can't distinguish mathematics from reality.

Starboy

[DigitalChicken]

Originally posted by schu
Anybody got the skinny on this guy? He has many of the marks of a fraud, but I'm not able to evaluate his work. He claims to have changed science as we know it. Any truth to that? Science News has an article this week that is only a bit skeptical.

You've got to read him and about him carefully and avoid fallacies when judging his work. Although I think it remains to be seen whether the ideas are worth their salt, I think a majority of the criticism of the work is unfair and much of it amounts to shooting the messanger rather than critisizing the message.

There are claims he is arrogant. That may be true but that doesn't speak of the truth of falsity of his ideas. Although he is somewhat bold and maybe arrogant he clearly recognizes that the ideas he is proposing are quite radical if true. He has to be given points for recognizing the analogy of his ideas to that of the invention of calculus. He makes this analogy and he is called arrogant for it.

His work is difficult to understand because it requires a level of mathematical sophistication. No doubt many of his critics have not digested his work. There were several condemnations very quickly after he released his "New Kind of Science" tome and there is NO WAY these critics fully or even adequately digested the ideas at that time.

Some say things about "he isnt saying anything new." This is often said with a tone that suggests he's trying to pul one over on the reader. However, he says rather openly that the basis for the ideas are not new. Part of the reason he wrote the book and took so long with it was, he claimed, because when he introduced some of these ideas previously in the early 80s they were quite brief technical and dry.

"A New Kind of Science" is often critisized because its not "new." However, that criticism fails to understand what is meant by the phrase "New Kind of Science." He makes an analogy to the impact and role of Newton's and Liebnitz's Calculus in science and especially that of how it evolved into differential calculus. To make a long story short a great amount of science that is based in mathematics is done with a dependence on modelling things in differential equations. In fact, when one models something new this is one strategy. In other words, one asks, "Is there a diffeq or a set of them that describe the phenomenon?" So in that sense he says the prescribed notion that one matches phenomenon to a an analytical equation or set of them is a "kind of science." He proposes a different method involving cellular automata which requires, in his opinion, a different set of mental procedures and a different attitude and mathematically one cannot use the same statistical methods that one used when using a differential equation. Thus, this different approach, which is radical, is a "New Kind of Science."

In the noted thread above, it appears CardinalMan, who is in this field, makes this error in not understanding what is meant by the title. This illustrates thats that some critics are either wrong, unfair, don't know what they are talking about, haven't digested the material well enough, or make some other error regarding it. Inf act CardinalMan admitted he only "skimmed" it. BTW, someone "in the same field" strikes me as odd since the methods that Wolfram proposes are not really used and thus its hard to claim that anyone is "in the field."

I think any honest and fair appraisal is yet to come because of the complexity of the ideas. Int he meantime one should avoid dismissing them for the wrong reasons.

DC

[Starboy]

Originally posted by DigitalChicken
I think any honest and fair appraisal is yet to come because of the complexity of the ideas. Int he meantime one should avoid dismissing them for the wrong reasons.

DC

DC did you read his book? It’s not very difficult to follow. His argument is as follows:

Lookie here, my automata can be manipulated to display shapes like you would find in nature such as snow flakes, cracks in crystals, fluid flow, complex behavior. I can create automata that emulate relativity, the motion of particles, nerve cells, logic and so forth. Therefore there must be something to this.

Well, there may indeed be something to it, but the only way for Wolfram to convince the scientific community is to start predicting phenomenon in nature that cannot be predicted by current science. Otherwise he is doing mathematics not science.

Automata are simple to understand. I found his book to be very approachable, if not annoying at times with having to wade through a lot of his self-aggrandizement. In some sense I hope he is on to something because if he were, it would make it possible for people like me to wade into the scientific fray again. I would like that very much.

Starboy

[DigitalChicken]

Originally posted by Starboy
DC did you read his book? It’s not very difficult to follow. His argument is as follows:

I haven't read that book completely but I have read portions of it. That would be obvious since a good portion of what I explained regarding the title was only understood by reading it.

However, I have read previous works including some published scientific papers of which the book is meant to expound. The papers are not necessarily easy to follow and require some workign through to understand them. I frankly don't think the book is as simple as you claim either but that's another story.

Well, there may indeed be something to it, but the only way for Wolfram to convince the scientific community is to start predicting phenomenon in nature that cannot be predicted by current science. Otherwise he is doing mathematics not science.

This is an unjustified statement of a philosophical nature and not a scientific one. There is no prerequisite in science that there can only be one explanation for phenomenon. Further, if you understood some of his examples then you would understand that he is actually, in a few cases, challenging current explanations.

DC

[Starboy]

Originally posted by DigitalChicken
This is an unjustified statement of a philosophical nature and not a scientific one. There is no prerequisite in science that there can only be one explanation for phenomenon. Further, if you understood some of his examples then you would understand that he is actually, in a few cases, challenging current explanations.

DC

Sorry, I never meant to imply that there could only be one explanation. But in the situation where there are two or more theories that explain the same things, it is the one that explains more phenomena with more accuracy that is usually favored. I do understand that he challenges some current explanations, but again, that is not the issue. Can he explain things that cannot be explained with current science? Otherwise all Wolfram has is a different representation for the parts of nature that are already explained by current science, interesting but unlikely to overturn the current thinking. This is all I meant by the comment. I actually hope he pulls it off even if he could use a few lessons in humility.

Starboy

[DigitalChicken]

Originally posted by Starboy
Sorry, I never meant to imply that there could only be one explanation. But in the situation where there are two or more theories that explain the same things, it is the one that explains more phenomena with more accuracy that is usually favored.

You and I both are getting off of the subject. :) However, your statement here is not entirely true as some explanations can in principle explain some aspects while not others. Think of a Venn diagram with where the circles represent the scope of explanation.

I do understand that he challenges some current explanations, but again, that is not the issue. Can he explain things that cannot be explained with current science?

It is the issue as science is in part about questioning theories in the light of new evidence and explanation. Also its not a matter of can "he" explain them. Its about the methodology.

Otherwise all Wolfram has is a different representation for the parts of nature that are already explained by current science, interesting but unlikely to overturn the current thinking.

This is wrong for two reasons. First, if the methodology can be made to explain things already explained by current science and the methodology explains it in ways for which there are no tranforms between the explanations then this is quite are revolutionary accomplishment.

The second error is that a few of the areas he challenges are "just so stories" such as proposed explanations for specific things in natural selection. If Wolfram is correct about even a single instance of this then he has either found an alternative framework other than natural selection or found parts of a model for its driving mechanism in whatever cases the example applies.

I think there are a few possibilities for the future of these ideas. One is that some bright mathematician(s) may be able to show that he is replicating results or that its coincidence. Another is that, if he is correct, then it might be a long time before these ideas are really applied usefully and as such we might never see it.

DC

[liquid]

I just tried writing out a fairly long post on this subject and then completely lost it! My first post for a while in fact, having been unwilling to access the site at work. I think I'm rusty!

I have read the entirety of the book, including the technical notes, although I'd be amazed if I remembered it all! As for my background, I'm studying aerospace engineering, which is primarily a physical, mathematical and computational degree. I have also spent a year as a research scientist, so I've got some context (though I would not say expertise) to judge the work on.

Basically, no, Wolfram is not a crank. He's a very clever mathematician who has made some important discoveries, particularly in the field of computationally tackling analytical maths (mathematica, one of the foremost maths software suites, was founded by him).

As is the case with many geniuses (and also cranks!) however, he is idiosyncratic. This is reflected in his approach to his work - he can be somewhat arrogant apparently. As mentioned previously, this has no bearing on his work fundamentally.

NKOS is essentially charting his research work, the work done previously to him, and puts forward his main hypothesis (the source of all the publicity hyperbole!).

It is true he does not credit in his book all work done by others. However, it has to be said that he does credit from time to time, and he certainly does in his academic papers. And again, a personally abrasive style doesn't really have any bearing on the work. However, the book is clearly written with excellent explanations of what can be a dry and complex subject, so don't I don't mean his writing style is poor - the presentation is just over the top.

Why?

1. It's a conscious reflection of Newton's Principia to publish as a book rather than through peer reviewed journals (which he abandoned after the first few years). This might be an ego/legacy thing, but it's not an invalid way of approaching things.

2. It gets people interested in reading about what can be a technical subject. I can't recall reading so much on one field of study I am not involved in; there's certainly more meat to it than something like A Brief History of Time.

3. It creates sales, but NKOS can't really be seen as a money making scam as it has taken 15 years or so of work!

So, presentation dealt with, what about his work? Well, he has made a few good discoveries of properties of cellular automata and related systems. This in itself constitues a reasonable body of work that would keep a mathematician in journals for a few years. Yet CA have never been seen as anything particularly relevant beyond the scope of maths and computing.

All the fuss is about Wolframs grand hypothesis, the principle of computational equivalance - and importantly, he does propose it as a hypothesis for research and testing, which despite the attitude problem pins it down as 'proper' science. A hypothesis is a relative of a claim, and needs defending, so I can see where Wolfram and his reviewers might get their wires crossed. It's just probably more tentative than Wolfram would like to think.

Basically, it states that once a system gets beyond a certain (very low) level of sophistication, is is as complex as anything else. This is based on observations of simple CA phenomena replicating far more complex phenomena, and of the discovery that there are whole swathes of mathematical systems that can be made to represent one another as soon as they get to this low level of sophistication.

We don't know whether he is right or not. Chances are he isn't as it stands, but we do have to recognise that informational/computational models of the universe are producing results, and are quite distinct from analytical models that we currently rely on if there is any such thing as traditional science (I'd rather not make the distinction!). After all, the universe can be considered as some kind of molecular computer/simulation. There are bound to be a hundred worthy Lamarcks before we get a new Darwin.

Finally, I'd like to comment on this;

Well, there may indeed be something to it, but the only way for Wolfram to convince the scientific community is to start predicting phenomenon in nature that cannot be predicted by current science. Otherwise he is doing mathematics not science

Please don't divide mathematics and science in quite that manner - the former is very much the language of the latter, and NKOS operates very much in the area of computer science, which is essentially pretty grey between the two fields. Wolfram's book does hypothesise plenty of predictions, and anyone who says otherwise simply hasn't read it thoroughly enough. The principle of computational equivalence certainly has predictive corollaries, some of which are explored. Of interest to me is a section predicting that CA system might have some ability to model turbulent flow, something we cannot satisfactorally do at present. He also predicts that network systems might be similar to a 'fundamental' universal system.

I think the best way to approach the book is as a hyped up but well written body of work, some of which is established, some of which is new, and then a bold hypothesis is appended and developed which is probably not correct on current knowledge but is worthy of investigation.

Incidentally, the one thing that really impressed me in the book is an explanation of how randomness might be generated by CA style phenomena. Traditional models of randomness are certainly severely lacking... ah you'll have to read the book, I'll shut up now!

[Jesse]

liquid:
All the fuss is about Wolframs grand hypothesis, the principle of computational equivalance - and importantly, he does propose it as a hypothesis for research and testing, which despite the attitude problem pins it down as 'proper' science. A hypothesis is a relative of a claim, and needs defending, so I can see where Wolfram and his reviewers might get their wires crossed. It's just probably more tentative than Wolfram would like to think.

Basically, it states that once a system gets beyond a certain (very low) level of sophistication, is is as complex as anything else. This is based on observations of simple CA phenomena replicating far more complex phenomena, and of the discovery that there are whole swathes of mathematical systems that can be made to represent one another as soon as they get to this low level of sophistication.

But as far as I can remember he never really defines the notion of "complexity" except in an intuitive way. Many might disagree that just because two systems exhibit computational universality, that necessarily means they are equally "complex". For example, see the distinction that this (http://www.goertzel.org/dynapsyc/2002/WolframReview.htm) review makes between "strong" and "weak" computational universality:

* Strong computational universality: All complex systems model classes that lead to nontrivial behaviors, lead to all possible complex behaviors, using reasonable space and time resources

* Weak computational universality: All complex systems model classes that lead to nontrivial behaviors, lead to all possible complex behaviors – but each model class only leads to a certain class of possible complex behaviors, under the restriction to use reasonable space and time resources

Neither of these has been proved to be true, and in fact, giving a sensible rigorous mathematical formalization for either of them would be a serious undertaking (a fact which Wolfram acknowledges, and attributes to the newness of the type of science he is doing).

Theme #3 of Wolfram’s book – that CA-ish systems are adequate for complex systems modeling in general – hinges on the strong computational universality principle. If only the weak computational universality principle is true, then it may be the case that some complex systems behaviors, while possible to simulate on CA-ish systems, are pragmatically simulable only via fundamentally different sorts of complex systems models.

Of course, the nasty little “efficiency” issue that separates my two variants of Wolfram’s principle, is also the biggest thorn in the side of the theory of universal computation (a standard part of theoretical computer science since Turing). Computation theory shows that any "universal computer" can simulate any other computer, by running an appropriate simulation program. It shows that some very simple computer programs are universal. But the catch is that, in practice, the simulation program may be unacceptably inefficient – the simulation of machine A, running on machine B, may run 100000000 times slower machine A itself, or may use 10000000000 times as much memory.

So, when Wolfram says that a given system – say, CA rule 110 – is “universal,” what this means is that it is in principle capable of modeling the universe, or the brain, or biological evolution, or whatever – if it is given a huge amount of memory and processing power (i.e. a huge initial condition and a huge number of iterations to run). It does not mean that CA rule 110 is in any way pragmatically effective for modeling all these systems.

If you have another modeling framework that you think is better than CA rule 110 for modeling, say, the brain or the universe -- then you are guaranteed that your whole modeling framework can be simulated using CA rule 110. But you’re not guaranteed that the CA rule 110 implementation of your modeling framework will fit into the observable universe, or run to the point of giving any meaningful result before the Big Crunch destroys us all.

[liquid]

Jesse, I actually agree (in the sense that I think they are valid queries) with most of those points expressed. I don't think you got the wrong end of the stick when replying to my post, but just in case - I'm not saying Wolfram is correct, just that the work is scientifically valid, in reference to the questions asked in the first post of the thread.

To address your/the reviewers points specifically, I would say this. Firstly that there are several competing definitions of complexity and all are in some way unsatisfactory - I cannot recall without the book to hand whether Wolfram makes absolutely no reference to definition issues; I suspect he does in fact, so that point must remain hanging until the proper digging has been done!

Secondly, the long passage about efficiencies of universal computation is all very well, but it misses the vital point that the discovery in principle is just as, if not more so, important than practical timescales of implementation. The reviewer seems to believe in this passage that the point of it all is to build a computer, not make scientific progress. (Note that I am commenting just on the quoted passage here)

The valid point the reviewer makes is actually here:

Neither of these has been proved to be true, and in fact, giving a sensible rigorous mathematical formalization for either of them would be a serious undertaking (a fact which Wolfram acknowledges, and attributes to the newness of the type of science he is doing).

Well, he's really answered his own criticism and admits Wolfram has too. As for addressing something as a serious undertaking, well, yes it is, but scientists can undertake things seriously.

[Jesse]

liquid:
Secondly, the long passage about efficiencies of universal computation is all very well, but it misses the vital point that the discovery in principle is just as, if not more so, important than practical timescales of implementation. The reviewer seems to believe in this passage that the point of it all is to build a computer, not make scientific progress. (Note that I am commenting just on the quoted passage here)

But it's not just a question of timescales, there's also the issue of what kind of behavior these systems will exhibit with "ordinary" initial conditions as opposed to ones that have been specially selected to get the systems to implement specific computations. If you look at rule 110's behavior with a bunch of different random initial conditions it'll probably look about the same on all the different runs, which seems a lot less "complex" then systems like brains or ecosystems which would exhibit qualitatively different outputs with different initial conditions.

liquid:
The valid point the reviewer makes is actually here:

Neither of these has been proved to be true, and in fact, giving a sensible rigorous mathematical formalization for either of them would be a serious undertaking (a fact which Wolfram acknowledges, and attributes to the newness of the type of science he is doing).

Well, he's really answered his own criticism and admits Wolfram has too. As for addressing something as a serious undertaking, well, yes it is, but scientists can undertake things seriously.

But is this undertaking anything that complexity theorists weren't already trying to do before Wolfram's book? If not, could you define what is the new problem that Wolfram has suggested?

[Starboy]

Originally posted by liquid
Please don't divide mathematics and science in quite that manner - the former is very much the language of the latter, and NKOS operates very much in the area of computer science, which is essentially pretty grey between the two fields. Wolfram's book does hypothesise plenty of predictions, and anyone who says otherwise simply hasn't read it thoroughly enough. The principle of computational equivalence certainly has predictive corollaries, some of which are explored. Of interest to me is a section predicting that CA system might have some ability to model turbulent flow, something we cannot satisfactorally do at present. He also predicts that network systems might be similar to a 'fundamental' universal system.

I think the best way to approach the book is as a hyped up but well written body of work, some of which is established, some of which is new, and then a bold hypothesis is appended and developed which is probably not correct on current knowledge but is worthy of investigation.

Incidentally, the one thing that really impressed me in the book is an explanation of how randomness might be generated by CA style phenomena. Traditional models of randomness are certainly severely lacking... ah you'll have to read the book, I'll shut up now!

Hi liquid, perhaps we will have to agree to disagree on this one. There is no doubt that mathematics is the primary language of science, however mathematics is definitely its own study and in its purest form wishes to have nothing to do with reality. You can do all the mathematics you want but until it intersects with reality it is only mathematics. Perhaps I missed something, but most of Wolfram’s hypotheses (speculations) were mainly about how automata could be used to explain phenomenon that already have explanations or how it might yield explanations for the unexplained but I didn’t actually see him make any reality predictions. If he could actually compute the rest masses of fundamental particles, or compute the mass of a neutrino or relate gravity to the other forces or any of the many unanswered question then everyone would sit up and pay a great deal of attention. That is all I was trying to say. That is not to say that if his approach can be shown to predict currently understood phenomenon that it wouldn’t be important, just not as important as Wolfram would like to claim.

As for Wolfram’s claim about generating randomness, well I find it a bit far fetched. All he has done is generate pseudo-randomness. Take an automata and use the same initial conditions and you get the same pseudo-random pattern. Even though the distribution of the automata generated pattern can be shown to be within the range that would be produced by a random pattern that would still not make it random. I don’t see how automata could be any better at predicting randomness than any other deterministic technique. When Wolfram gets to QM he sort of hand waves over this difficulty. He is very quick to demonstrate that his automata are intrinsically quantized but glosses over the very deterministic generation of the actual patterns. His primary problem is that a real random pattern would not be computable. The best he could do is compute all possible patterns and take a classical Boltzman ensemble approach, but again, nothing new there.

Starboy

[fando]

I feel that all the negative fuss about Wolfram's ANKoS is premature and pointless. Most of the time, he is flamed for not going through the normal peer review channels, for being an outside player. For instance, in the article Jesse posted, there is nothing but ad hominem attack on Wolfram for his lack of attributing credit and a trivialization of his presentation as nothing new. Those are very valid criticism... for a research paper that is going through the formal peer review process. However, Wolfram isn't publishing a research paper or a formal textbook here, just a popular science book of a different kind, one meant to provoke discussion about a certain subject. So who cares if he doesn't attribute credit and doesn't go through a journal first? I didn't give a damn myself, I was just interested in the central thrust of his work. And who cares if his presentation is egoistical? That he's pissing people off only seems to serve his purpose, which is to bring attention to what he had in mind for ANKoS.

So, what did Wolfram have in mind? He seems to be proposing that we think more about information, computation and how it ties in with reality. Wheeler is now doing work with the information entropy of black holes and where information and computation fit into the grander scheme of things. I think that's exactly what Wolfram wants to see more of: Talk about information, computation and reality. It's as simple as that! And we don't have to revolve around Wolfram's ideas--he is merely providing some initial momentum by showcasing his own work. We're free to criticize what he puts on the table, but there's no good reason to ignore the spirit of his effort and not provide alternative ideas along the same vein. Don't like his principle of computational equivalence? Come up with something better! I recall Wolfram stating these points multiple times throught the book, but almost everyone seems to overlook them when talking about ANKoS. Instead, he is attacked on the specifics and because of a weakness in one or another, the whole work is dismissed. The book is not really about trees, but about a forest.

[liquid]

But it's not just a question of timescales, there's also the issue of what kind of behavior these systems will exhibit with "ordinary" initial conditions as opposed to ones that have been specially selected to get the systems to implement specific computations. If you look at rule 110's behavior with a bunch of different random initial conditions it'll probably look about the same on all the different runs, which seems a lot less "complex" then systems like brains or ecosystems which would exhibit qualitatively different outputs with different initial conditions.

That's all true, again it's a point that needs addressing. But pointing out 'issues' with hypotheses isn't a problem; perhaps you are expecting it would be. If so, you misunderstand what I'm trying to get across; I spent a large part of my post explaining that there may well be too many issues, but you have to investigate to find out, rather than just mooting the possibility and leaving the critique at that.

But is this undertaking anything that complexity theorists weren't already trying to do before Wolfram's book? If not, could you define what is the new problem that Wolfram has suggested?

1. Yes.
2. Firstly, my point relates to the review (or rather the quoted problem) and the reviewer. So the second question neither follows directly or has much relevance to the point I was making, and I'm having a little trouble understanding the nuance without a clear idea of context (that's probably not a clear sentence in itself!).

As far as I understand the question, my answer would be Wolfram isn't redefining the universal computation problem - nor has he claimed to, which makes the question seem odd to me, but suggesting a framework in which the answer may be found, and proposing his own speculative hypothesis.

------------------------------------------------------------------------------

There is no doubt that mathematics is the primary language of science, however mathematics is definitely its own study and in its purest form wishes to have nothing to do with reality

This gets us onto tricky philosophical ground! I could start pointing out that all mathematics, despite pretensions to abstraction, is all performed within reality. However, it just turns out that my complaint was not about distinguishing maths and science, but just the manner in which it was done. This is because the study of CA immediately impacts upon reality, if by no more than the fact that they are computed by physical systems (i.e. computers!). Perhaps I wasn't clear enough that they depend very much on the physical computational process and are therefore as much a part of computer science as mathematica code. So I don't disagree with the original statement so much as its application to this work.

Perhaps I missed something, but most of Wolfram’s hypotheses (speculations) were mainly about how automata could be used to explain phenomenon that already have explanations or how it might yield explanations for the unexplained but I didn’t actually see him make any reality predictions

Hypotheses are speculations by any other name, just scientifically framed. And many physical systems have several explanations, that does not make new ones invalid. Remember we are talking scientific models here, not fundamental reality.

As for 'reality' predictions, I've already listed some. Perhaps I should be more specific, and point out the types network patterns that Wolfram thinks might be able to represent atoms, or the CA type process the thinks is behind real animal patterning. Sorry, but this annoys me because there just are predictions available, although it would be nice if quantifiable masses of electrons could be given, I think you will have to understand that it's not at the level of development to be able to give such numbers.

I totally agree that such problems WILL need to be solved for the 'sit up and take notice' effect, but the inability to do things like that does not disqualify it as valid scientific proposal - 99.99% of all the papers published don't even get near such paradigm shifting discoveries. It's perhaps unfortunate that Wolfram chose to claim a paradigm shift, but this should never be about defending or attacking the man.

Going to have to stop now, so I'll address the rest in a bit.

--------------------------------------------------------------------------

Quick PS: fando, just caught your post when relogging in to post this. Interesting view to take.