 |
Author |
Topic: Wolframs New Kind of Science... (Read 640 times) |
|
jand
 
|
Wolframs New Kind of Science...
« on: Mar 22nd, 2003, 9:10am » |
|
I haven't seen this mentioned here, which surpises me; then again, I'm a noob so apologies if this has already been discussed...
It's pretty hard going in places; I don't come from an academic background but Alife/System theory has always been a main area of interest, so I didn't find it insurmoutable...
I'm about to have a re-read and this time through I think P5 will make a great companion...
So has anyone else read it or used any of the ideas therein with their P5 work...collabs?...
|
| « Last Edit: Mar 22nd, 2003, 9:11am by jand » |
|
|
|
|
Martin
   
|
Re: Wolframs New Kind of Science...
« Reply #1 on: Mar 22nd, 2003, 11:28am » |
|
a lot of it deals with complexity theory and cellular automata. there are quite a number of p5 pieces that show cellular automata. for instance, mike davis has a rather fantastic collection of these. glen murphy's fluid dynamics stuff is nice. the book doesn't really present anything new. people have been doing these stuff for quite some time now... even in the 70's (refer to conway's game of life). i quote from amazon.com: mega biblia mega kaka
hmm... a phd at 20 is still nice. 
trivia: 1950s is considered the period when cellular automata emerged
at any rate, it's a good exercise 
|
| « Last Edit: Mar 22nd, 2003, 11:35am by Martin » |
|
|
|
|
benelek
|
Re: Wolframs New Kind of Science...
« Reply #2 on: Mar 22nd, 2003, 4:44pm » |
|
on Mar 22nd, 2003, 11:28am, Martin wrote:| trivia: 1950s is considered the period when cellular automata emerged |
|
haha, i think cellular automata has been around a lot longer than that. just ask a biologist/anthropologist
of course, there weren't any trendy magazines to publish the inner workings of them at the time!
|
|
|
|
|
jand
|
Re: Wolframs New Kind of Science...
« Reply #3 on: Mar 22nd, 2003, 5:38pm » |
|
I think you're dismissing the scope of his work considerably by limiting it to CAs...
Sure, he uses them as an easily explained stepping stone onto far more wide-reaching areas...
By formulating science as "small programs" rather than Maths equations his approach seem almost custom built for exploration in P5 etc...
My suggestion is to grab a copy..(develop those muscles first.....and have an explore - I think you'll be surpised how far he's able to take his approach...
|
| « Last Edit: Mar 23rd, 2003, 1:32am by jand » |
|
|
|
|
Martin
|
Re: Wolframs New Kind of Science...
« Reply #4 on: Mar 23rd, 2003, 4:28am » |
|
i have it. well, not really mine but have access to one. hmmm, i guess ca is the dominant thing in his book. ... re: formulating science as small programs... hmm... "buy mathematica"? ... the idea of formulating science as small programs is quite old already. however, it's not a reason not to pursue them.
|
|
|
|
|
Martin
|
Re: Wolframs New Kind of Science...
« Reply #5 on: Mar 23rd, 2003, 4:30am » |
|
on Mar 22nd, 2003, 4:44pm, benelek wrote:| haha, i think cellular automata has been around a lot longer than that. just ask a biologist/anthropologist |
|
however, they might not have enough computing at that time. first generation computers appeared arnd 1951.
|
| « Last Edit: Mar 23rd, 2003, 4:31am by Martin » |
|
|
|
|
jand
|
Re: Wolframs New Kind of Science...
« Reply #6 on: Mar 23rd, 2003, 9:00am » |
|
"quite old"?...perhaps....
Largly unexplored though, due to the sheer processing power needed I suspect...and if Wolframs systematic expolration is anything to go by, worthy of far more serious attention...
I think people have been sidetracked by the CA bias; they are simply a device (a notation??) rather than the subject itself...
p.s. ...wish I could afford Mathmatica...p5 will have to do for my Yob Maths.....
|
|
|
|
|
benelek
|
Re: Wolframs New Kind of Science...
« Reply #8 on: Mar 25th, 2003, 12:47pm » |
|
on Mar 23rd, 2003, 4:30am, Martin wrote:
however, they might not have enough computing at that time. first generation computers appeared arnd 1951.
|
|
ah, but i dont just mean in computing.
ever since the rise of euclidean geometry (beginning much earlier, probably around the time of Pythagoras), there has been what i like to think of as Modernism. it's a way of thinking that produces the defined patterns as an operable set, such as equations. this way of thinking has grown hugely in the last few thousand years, in the last few hundred years manifesting as science/technology. it allows the evolution of predictive methods that put beginings and ends of a system in a common frame of reference. this approach is the fundamental opposite of CA.
i'm afraid i'm being a little vague here, but i'll keep going...
people tend to compare the spiral-like shape of a sunflower to the golden ratio in geometrical form. many of my fellow students are amazed that a flower could be "intelligent" enough to follow the shape of a complex mathematical form, and it seems to me that we have begun thinking in a backwards manner, leading from solution to method - a somewhat non-darwinian approach. we can predefine the shape of a flower in a computing package by using the mathematics of the golden ratio. but we don't know how to approximate the golden ratio by simulating evolution.
would i be right in saying that cellular automata refers to the pull-and-tug that occurs in the process of evolution to produce forms that approach an ideal form through the contant influence of, and reaction to, certain forces
take, for instance, the generation of a parabola on screen.
one way to do it would be to provide the equation and increment along the x-axis, feeding the points in to the equation in order to provide the y-value.
another way would be to have one float for the y-position, one float for the rate of change of the y-position (velocity, first derivative), and one float for the increase/decrease in change-of position (acceleration, 2nd derivative). you would scan along the x-axis, increasing the y-position by the velocity and increasing the velocity by the acceleration at each increment along the x-axis. ie,
Code:
color thePix = color(0,0,0);
float yAccel = -9.8;
float yVel = 0;
float yPos = height;
for(int x=0; x<width; x++) {
if(yPos>height) {
setPixel(x,yPos, thePix);
yVel+=yAccel;
yPos+=yVel;
}
} |
|
the above code would draw a parabola, using concepts of physics more than concepts of mathematics. as it happens, i've simulated a falling object under gravity. note that this method approximates the form of a parabola without knowing what the equation of one. this method is much more flexible, because it doesn't require the knowledge of any equation. it works through quantasizing complex forces into simple increments.
-jacob
|
| « Last Edit: Mar 25th, 2003, 12:48pm by benelek » |
|
|
|
|
jand
|
Re: Wolframs New Kind of Science...
« Reply #9 on: Mar 26th, 2003, 12:19am » |
|
good post....early here but will reply more in the morning...you bring up some interesting points.....
but "but we don't know how to approximate the golden ratio by simulating evolution. "...
yes, we can....137.5 (the golden angle) evolved due to the fact that's it's an optimum packing method for petals etc.....
|
|
|
|
|
benelek
|
Re: Wolframs New Kind of Science...
« Reply #10 on: Mar 26th, 2003, 1:05pm » |
|
mm, i was being vague. the global "we" (the authoritative science) certainly knows the connection
but i'd challenge someone to come up with an applet that can (through the use of simple CA methods, and without the golden number) generate the optimumly packed petal! extra points for someone who does it by also using CA methods to generate the reasons for a flower to be so closely packed... a cause-and-effect sketch? lol.
as an unrelated piece of conversation - there was a student Rally Against War here today which turned into a Riot Against War. fighting war with violence... like fucking for virginity, ey?
|
|
|
|
|
|
