To the mods I am posting this here because of the debatable nature of the import of this self published book. If it needs to be moved to Cafe Society please do so.
Has anyone read this new book by Stephen Wolfram, the inventor of Mathematica? It’s supposed to turn science on it’s ear. Freeman Dyson thinks it’s thesis is wildly overstated re the “end all, be all” import the book is claiming for the notion that the universe is defined and determined by the laws governing the action of cellular automata.
I haven’t read it (who has? It’s 1000+ pages) but it doesn’t sound so revolutionary. People have been doing research on self-organizing systems for many, many years. What is so revolutionary about a book about an established field of study?
I have never heard of cellular automata, but I did read a little bit about complexity theory in a class on organizational behavior. Since chaos theory implies that the outcome of a system is unpredictable even when you have a complete set of initial conditions, I wonder if those few simple rules would be meaningful in an applied problem. Even if you can identify the Pekingnese butterfly that contributed to last month’s hurricane, you won’t be any better off in trying to predict hurricanes next month.
OTOH, I do remember from the review that Wolfram’s book did call into question some key assumptions in science, such as the second law of thermodynamics (order inevitably disintegrates into disorder) and the importance of natural selection in explaining the variety of species we see in nature.
I think you’re going to get a lot of WAGs in this thread, by people who don’t really know what the book is about.
However, the general feeling I got from the physics department in Urbana, IL (the town where Wolfram Research is based), is that Wolfram is off his rocker. When I graduated, the book still hadn’t come out, but some people had read exerpts or preprints. Now that it’s out, I’ll have to go to the bookstore and flip through it, so I can report back more directly.
As a general rule, any time someone publishes something that will “turn science on its head” without going through the peer review process, it’s usually crap.
As another rule, when a scientist leaves his field of study for a wildly different one and starts making claims about overturning that whole area (as Wolfram has apparently done here, from what I’ve heard), it’s also usually crap. (For instance, Linus Pauling and Vitamin C.)
Just because system A produces results that look like system B, does not mean that system B is governed by rules similar to the rules governing system A.
For example, a spiral galaxy bears a remarkable resemblance to a hurricane. However, the gravitational mechanics that create the spiral galaxy shape are almost nothing like the air-mass dynamics that create a hurricane. The resemblance is entirely superficial. (Wilhelm Reich fell into precisely this trap, by assuming that because hurricanes and galaxies looked the same, they must be governed by the same “common functioning principle” – which Reich, of course, assumed was orgone energy. Reich thought everything was orgone energy.)
Similarly, an algorithm that generates images which look like a leopard’s spots may have nothing in common with the genetic/enzyme program inside the leopard that created its spots.
If that really does represent Wolfram’s understanding of the second law of thermodynamics, we’re in trouble. On the other hand, if that really does represent your understanding of the second law of thermodynamics, you’re in trouble. Which is it? Or is it neither, simply an overly succinct statement of the second law?
Every time I hear about something turning the establishment on it’s ear, I wait a few years. Haven’t seen any establishments on their ears yet, don’t think I will any time in the near future.
Well, Wolfram put cellular automata on the map. So, that may make it potentially quite revolutionary coming from the source so to speak (then again, Einstein put relativity on the map and look what happened to his unified theory search.)
David B offered:
I don’t disagree with this statement so long as the caveat “leaves his field of study for a wildly different one” applies. But, I’m not sure that Pauling/vitamin C are analogous to Wolfram/order. Pauling had no business speculating about medicine. But, Wolfram speaks with authority about emergent order.
I just received it, and I’m into the third chapter where he starts making his point about cellular automata with some demonstrations. They are surprising.
Wolfram’s point isn’t that cellular automata (or something like it) are at the root of everything. His point is that modern science is built upon math, insofar as it tries to reduce its objects of study to equations. That practice has been very successful in physics, but it’s hitting walls where equations are becoming more complicated, less descriptive, less accurate, and generally less useful in generalizing phenomena. In other fields, like social sciences, is nearly useless.
Wolfram’s idea is that a generalization of math–computation–is the next big step to take. Rather than reducing complex phenomena/observed behavior to equations, they can be reduced to simple algorithms. Do loops, rather than equals signs.
Cellular automata are the starting point because very simple rules can generate very complex patterns, which demonstrates that complexity can easily arise from very simple, iterative programs. But they aren’t the entirety of his theory.
I have no clue whether it will still make sense by page 800, but it’s not tinfoil hat material, by any means.
Cellular automotons! Of course! How could I miss that! Even if you prevent the broadcasts from reaching your dental fillings, they can still breach your defenses using tiny atomic robots!
As I stated before–rather clearly, I thought–I am repeating the interpretation given in the review. We’re having a polite conversation here, ultrafilter, and there’s no need to get pissy in your first contribution to the thread.
BTW, this page sponsored by Berkeley’s college of chemistry states:
Duhhh, I’m just a drooling nonscientist . . . duhh, I shouldn’t even be posting in this thread for fear that I’ll short out my keyboard through my slack-jawed drooling . . . but isn’t the statement “order disintegrates into disorder” sort of a workable thumbnail sketch of this definition? I suppose your quibble is with the word “inevitably”, which as I said, was the reviewer’s word and not mine.
My understanding of complexity theory is from the perspective of organizational behavior, not physics. Nonetheless, the assertion that entropy is not inevitable appears consistent with what little I’ve learned so far about self-organizing behavior.
Okay, this GD, not Cafe Society. So rather than worry about what he actually does say (like others here, I’ve read reviews but it hasn’t made the top of my reading list) how about debating the value of what it is portrayed as saying, which seems to boil down to this-
Mathematics as traditionally considered is limited in the analysis of complex systems operating in real time. Computer modelling of multiple agents operating according to fairly simple algorithms may be more practical. With a lot of pretentious self aggrandizing crap added in.
Sounds like he is repackaging the fields of nonlinear dynamics and of multi-agent computing, both well established areas, and calling it all his new idea.
Chaos theory, BTW, is a lot more than stating that “the outcome of a system is unpredictable even when you have a complete set of initial conditions”. It recognizes that in complex nonlinear systems (most of the real world) minute changes in intial conditions can have dramatic effects on the final results. Direct calculation of the final results from initial conditions is undoable. But the results form particular sets of solutions that are predictable, that form into patterns. Let the complex nonlinear system run and it will settle into one of these “attractor basins.” The patterns of these sets of solutions are sometimes fractal sets that are often self-similar at different levels of analysis.
So, sure, that is how the real world operates. In astrophysics where the effects of multiple bodies on each other and back again simultaneously must be considered; in quantum physics; in neurobiological systems; in social structures; in the formation of epistemologies themselves. At all levels. And the way to model it is with parallel computing with multiple agents and with a solid understanding of nonlinear dynamics in all its complexity. Maybe not fully appreciated yet, but not a new idea.
BTW, why does this book get reviewed in all the business sections?
As long as I’m here, I guess I might as well explain my issue with that statement of the second law. To me, that makes it sound like order must always decay into disorder, with no room for a local increase in order. This statement is false, but creationists love it, and so it always makes me a little nervous to hear stuff like that.
A better soundbite is “A closed system will experience a positive change in global entropy as time goes forward”. If the system isn’t closed (like the earth, which gets a lot of energy from the sun), there doesn’t have to be an increase in global entropy. And you can have a local decrease in entropy (i.e., a spontaneous increase in order) as long as the overall entropy of the system increases. For instance, even though I’m increasing the amount of information here by typing these words, the energy lost as heat by my body while I type more than compensates for it, and then there’s the energy required to display them on screen, and transmit them, etc., etc. I hope that’s clear.
Anyway, I don’t want to hijack this into any other territory. Wolfram’s book is probably one of those things that I’ll read someday, but I’ve got a lot of other stuff to get through first.
As others have said before. His ideas are either new nor revolutionary.
Von Neumann was toying around with cellular automata before Wolfram was even born (I might be getting the time scales a bit wrong but von neumann did his work in the 40’s/50’s and Wolfram is ~40 years old right).
Computation has been PART of mathematic for donkeys years in maths terms and computability and recurrence/recursion has been one of the many thriving areas of maths. Non-linear systems used for modeling have been prevelant in Economics for donkeys years as well and the fact that cellular automata looks like stuff is hardly suprising news.
In fact, I have a copy of New Scientist from the early 1990’s where some guys found that defects in a cellular automata grid actually ated like fundamental particles with spin and mass.
The big deciding factor for me was when wolfram did an interview where he was basically pimping himself in the most base manner. Clearly he has no idea about the scientific process and, when, if ever, he produced a paper in a peer reviewed journal, THEN I will take him seriously.