5 ms·
I haven't read the book, but I've dipped into it. Some of the ideas are interesting; ideas that I was acquainted with through Wolfram's papers before the book a
by leephillips 10y ago
I haven't read the book, but I've dipped into it. Some of the ideas are interesting; ideas that I was acquainted with through Wolfram's papers before the book appeared. From critical examinations published by others, I've gleaned two things: the book contains one new theorem, apparently discovered by one of Wolfram's collaborators; and, there are no successful predictions in the book. To the extent that it's supposed to be "science", 15 years later one might hope to see the author saying something like, "Here is my NKS prediction of <whatever>, and this experiment confirms it," or "this experiment contradicts my prediction." Even the latter case would show an attempt to do science. So I'm intensely interested in knowing, from people who have been following this, are there any testable predictions from this NKS, have the experiments been done, and what were the results? Because this new article (I couldn't get through it, I confess) seems to be just more handwaving, and Wolfram talking about how important his insights just have to be.
EDIT: I meant to add that it's very nice that the book is now free, in any case.
- cr0sh 10y agoMaybe the choice of title was a bad move on Wolfram's part? Your argument makes sense; it doesn't predict anything, or show any hint of science. Based on what I recall of the book (I did read it - everything except the appendix/notes), and what I've read of this new article (I didn't finish it - I will read it in full at a later point, though), it can't predict anything. What I mean is that these simple rule programs can't be run any faster than the universe in which they are running. So you can't say "I predict this simple rule set will eventually draw a butterfly" (or whatever) - because you couldn't run it long enough for that to occur. In a way, it's a simulation (or emulation) problem; just as you need a more powerful computer to run software to emulate a system in real time, provided that system is less powerful - you would need to do the same in order to run these simple programs. In essence, since the theory espoused in NKS (more or less - at least, that is what I got out of it - and maybe I am wrong) is that the universe at its base level (quarks? dunno) is composed and operates and creates based on these simple rules, in order to go faster than our universe (and catch up), you'd have to have a more powerful system than whatever the universe is running on (conjecture, I know - and Wolfram noted this as well IIRC). So right now - just like if you were wanting to emulate a PS4 on a contemporary PC - the only way you can do it is non-realtime; you have to be satisfied at running the system at a slower rate. What is interesting about this - and this is pure speculation - let's say you are running "rule 30" on a computer, and unknown to you, it is actually computing a new universe (that is, simulating it) - to you, it is happening much more slowly than realtime - but to anyone "inside" that universe, they can't relatively know that they are running slower - their sense of "time" is "normal"... I'd like to encourage you to read the book in full and maybe this new article as well. Yes - it's a real long slog. And maybe it is nothing more than a giant ego stroke on the part of Wolfram; I don't really know what to think about it anymore. But you might walk away (to get some aspirin for your headache?) with something to think about and ponder - even if it is completely wrong, or you think so, or whatever - simply reading it and understanding it might lead you to your own discoveries and ideas.
- leephillips 10y agoThat's an interesting point of view. I've never heard this work described that way before. If that's the correct interpretation, though, (and I think you've implicitly acknowledged this), the NKS work is, perhaps, some kind of metaphysical speculation, with some demonstrations to support it -- but not any kind of science, since, as you say, it can't predict anything. Also, in that case, it would be unfair to demand that it produce a prediction, of course.
- xrange 10y ago>I've never heard this work described that way before. One way to summarize the thesis of book would be to say: Historically, we've analyzed natural phenomena and tried to fit equations to the data. But Nature may be doing something analagous to running simple programs instead of solving equations. What are the implications of this? Complexity and "randomness" comes about from the "programs", not from their initial inputs/conditions. And predicting behavior of programs is harder than predicting behavior of systems of equations, which sucks. >but not any kind of science I think many people who aren't familiar with the book are thinking that "a new kind of science" means "a new branch of science" like we have physics, chemistry, biology, and now this new "computery" science. But the book is actually about looking at nature with the new perspective of "programs", that previous generations of people didn't have a reference frame for. And some of things that we think are "complex" might actually be following simple rules (or can better be modeled as such).