John
More interesting here
It's quite interesting to see the general makeup of the Orkut community. Most who are active can click on just a few friend links and quickly feel like 'yeah, i know most of these people'.
What will be interesting is how this will change as Orkut grows in popularity and, possibly, becomes much more homogeneous.
Right now it seems to be made up, primarily, of atheists and game developers.
So, here's my question. I am an atheist, like many here. I'm also a big believer in evolution as well as artificial life research and emergent behaviour in complex systems.
However, I'm also a big believer in the 'put up or shut up' philosophy.
Most of us in game development have a very strong engineering attitude. We make shit work, and while we respect the ideas that come out of academia we have little respect for their work ethic.
That lengthy preamble aside, I have a problem with academia and artifiial life. I see an awful lot of talk and very little action.
The premise is put forth. It sounds good but the actual implementation is missing! A few weak demos that explore the boundaries of mathematical space do not a truly complex system make.
To date must demonstration applications of evolutionary based models in action amount to little more than a heuristic applied to culling mathematical space. Sometimes you get some decent looking demos, but evolving something that feels truly original or 'new' is really lacking.
I'm waiting to actually see a demo that experiences puncuated equilibrium in the sense that I am shocked and amazed when I seem to get more information out of the system than was put into it.
There is this idea of the law of thermodynamics where you can't get more energy out of a system than you put into it. And, that rule holds true quite well. However, equating energy with complexity is not quite so simple.
In most cases when energy is injected into a system the end result is less information not more. In short, you blow shit up and you don't expect DNA!
What will be interesting is how this will change as Orkut grows in popularity and, possibly, becomes much more homogeneous.
Right now it seems to be made up, primarily, of atheists and game developers.
So, here's my question. I am an atheist, like many here. I'm also a big believer in evolution as well as artificial life research and emergent behaviour in complex systems.
However, I'm also a big believer in the 'put up or shut up' philosophy.
Most of us in game development have a very strong engineering attitude. We make shit work, and while we respect the ideas that come out of academia we have little respect for their work ethic.
That lengthy preamble aside, I have a problem with academia and artifiial life. I see an awful lot of talk and very little action.
The premise is put forth. It sounds good but the actual implementation is missing! A few weak demos that explore the boundaries of mathematical space do not a truly complex system make.
To date must demonstration applications of evolutionary based models in action amount to little more than a heuristic applied to culling mathematical space. Sometimes you get some decent looking demos, but evolving something that feels truly original or 'new' is really lacking.
I'm waiting to actually see a demo that experiences puncuated equilibrium in the sense that I am shocked and amazed when I seem to get more information out of the system than was put into it.
There is this idea of the law of thermodynamics where you can't get more energy out of a system than you put into it. And, that rule holds true quite well. However, equating energy with complexity is not quite so simple.
In most cases when energy is injected into a system the end result is less information not more. In short, you blow shit up and you don't expect DNA!

12/03/2004
John
continued
...so that's really my point here. When you read the published works in academia on artificial life research it comes off with a a very pedantic tone implying that emergent behaviour 'just happens' with very little effort or energy.
However, to date, no one has actually produced any compelling results. A few flocking boid demos just don't cut it. Demos which apply a human created heuristic to explore the extent of mathematical space don't cut it either.
So, what is the problem? Is the premise flawed or is the problem that the simulations don't operate in an environment that is sufficiently rich to approach the complexity of the natural world?
If we have more powerful simluation tools will our artificial life simulations suddenly blossum? Or, will we only be exploring richer mathematics but not achieving anything akin to the 'spark of life'?
How does a giant explosion, the big bang, lead to complexity of information? Normally when I blow shit up it becomes less complex, not the other way around?
Is there even a mystery here at all?
I throw this challenge down. No artificial life researcher is allowed to write one more paper about how 'complex behaviour naturally arises from a set of simple rules' without reproducing something of equal complexity from similar initial condtions.
The simple demos seem to produce simple results. The complex demos seem to reflect the complexity of the software that was invested into by the by creative mind of the author rather than anything uniquely produced by the system itself.
As an atheist I am fundamentally a skeptic at heart, and colour me skeptical of these Alife researchers until they 'put up or shut up.'
However, to date, no one has actually produced any compelling results. A few flocking boid demos just don't cut it. Demos which apply a human created heuristic to explore the extent of mathematical space don't cut it either.
So, what is the problem? Is the premise flawed or is the problem that the simulations don't operate in an environment that is sufficiently rich to approach the complexity of the natural world?
If we have more powerful simluation tools will our artificial life simulations suddenly blossum? Or, will we only be exploring richer mathematics but not achieving anything akin to the 'spark of life'?
How does a giant explosion, the big bang, lead to complexity of information? Normally when I blow shit up it becomes less complex, not the other way around?
Is there even a mystery here at all?
I throw this challenge down. No artificial life researcher is allowed to write one more paper about how 'complex behaviour naturally arises from a set of simple rules' without reproducing something of equal complexity from similar initial condtions.
The simple demos seem to produce simple results. The complex demos seem to reflect the complexity of the software that was invested into by the by creative mind of the author rather than anything uniquely produced by the system itself.
As an atheist I am fundamentally a skeptic at heart, and colour me skeptical of these Alife researchers until they 'put up or shut up.'

12/03/2004
John
missing the point...
the premise from acedimia is that it's "all quite simple". "Simple rules produce complex behaviour". "Poof".
If it's so freaking simple, why can't they produce something other than a paper to pad their CV?
It's a reasonable question.
As I said, I actually believe the premise has merit becuase I believe the universe operates this way. I believe the universe operates this way because it meets my observations and personal experience.
The post was relevant to the discusion of atheism in this sense.
You can't get something from nothing.
You can't blow shit up and get DNA. It doesn't work that way.
The sum is more than it's parts.
The universe is and of itself a highly, highly, complex system. I think that's rather obvious since you exist within it and bother to question this bad boy. The Universe is god and is clearly responsible for everything that is, has been, and ever will be.
Why a researcher can't write a piece of software that can produce a decent looking anthill is beyond me.
If it's so freaking simple, why can't they produce something other than a paper to pad their CV?
It's a reasonable question.
As I said, I actually believe the premise has merit becuase I believe the universe operates this way. I believe the universe operates this way because it meets my observations and personal experience.
The post was relevant to the discusion of atheism in this sense.
You can't get something from nothing.
You can't blow shit up and get DNA. It doesn't work that way.
The sum is more than it's parts.
The universe is and of itself a highly, highly, complex system. I think that's rather obvious since you exist within it and bother to question this bad boy. The Universe is god and is clearly responsible for everything that is, has been, and ever will be.
Why a researcher can't write a piece of software that can produce a decent looking anthill is beyond me.

12/03/2004
Oran
In defense of artificial life
I guess I'm an engineer who wants to be an academic, but I also want money now. Maybe I don't know the lit as well as you, but....
Have you seen avida? Perhaps it doesn't quite live up to your standard of getting a lot of complexity from a very simple system, but it’s interesting. It's an alife program with a virtual world of simple computers with a simple machine language. Organisms are programs that can propagate themselves, and programs are killed periodically for being unfit. Some things that are counted towards fitness are (a) how well they compute some reasonably simple function(s) and (b) how short the code is. The funny thing is, if you ask a programmer to come up with the shortest possible program to compute these function(s), it turns out that avida can produce shorter programs than programmers! OK, it's really not a practical tool for programmers. But it does let us look into the "physics" of life and evolution, if there are such physical laws.
And it turns out that what you might call punctuated equilibrium happens in the form of rare but revolutionary changes that rapidly spread and replace any programs without it. During the early phases, there is a rapid growth in diversity, but as the new approach is perfected by incremental improvements of small mutations, the population eventually comes to an equilibrium again... until bam, a scientist invents something that obsoletes the dominant engineering framework... er, what was I saying?
....There are some practical IP packet routing strategies in "swarm intelligence" using ant-like behavior....
Also, IMHO, boids are not artificial life -- not even "swarm intelligence" -- because they don't really try to do anything substantial that a slightly smarter boid couldn't do. What they are is a model of flocking behavior that may or may not be useful for behavioral biologists or game developers.
Have you seen avida? Perhaps it doesn't quite live up to your standard of getting a lot of complexity from a very simple system, but it’s interesting. It's an alife program with a virtual world of simple computers with a simple machine language. Organisms are programs that can propagate themselves, and programs are killed periodically for being unfit. Some things that are counted towards fitness are (a) how well they compute some reasonably simple function(s) and (b) how short the code is. The funny thing is, if you ask a programmer to come up with the shortest possible program to compute these function(s), it turns out that avida can produce shorter programs than programmers! OK, it's really not a practical tool for programmers. But it does let us look into the "physics" of life and evolution, if there are such physical laws.
And it turns out that what you might call punctuated equilibrium happens in the form of rare but revolutionary changes that rapidly spread and replace any programs without it. During the early phases, there is a rapid growth in diversity, but as the new approach is perfected by incremental improvements of small mutations, the population eventually comes to an equilibrium again... until bam, a scientist invents something that obsoletes the dominant engineering framework... er, what was I saying?
....There are some practical IP packet routing strategies in "swarm intelligence" using ant-like behavior....
Also, IMHO, boids are not artificial life -- not even "swarm intelligence" -- because they don't really try to do anything substantial that a slightly smarter boid couldn't do. What they are is a model of flocking behavior that may or may not be useful for behavioral biologists or game developers.

12/03/2004
> "Simple rules produce complex behaviour"
I've found that to be quite true.
But 99% of the time, the "complex behavior" is a bug.
I've found that to be quite true.
But 99% of the time, the "complex behavior" is a bug.

12/03/2004
St. Chris
I'm still puzzled why this discussion isn't in, say, the Artificial Life forum -- but then, I don't hang out there, so it's cool to get a little of it dropped in my lap.
John wrote:
the premise from acedimia is that it's "all quite simple". "Simple rules produce complex behaviour". "Poof".
Well, yeah, if eleven billion years is "poof" to you. Or are the premises in a-life research irrelevant to physical science? Am I missing some nuance because I'm not in the field?
John again:
You can't get something from nothing.
You can't blow shit up and get DNA. It doesn't work that way.
Big Bang -> us. No?
John once more:
The Universe is god and is clearly responsible for everything that is, has been, and ever will be.
Waitwaitwait. Can you clarify your use of "god" there? Do you mean simply that the Universe is self-originated and self-governing? If you apply the g-word in that case, you seem to be advocating scientific pantheism -- which is a very nice philosophy, but isn't atheism.
I am interested in what you're saying, John. I just want to make sure I understand your terms.
John wrote:
the premise from acedimia is that it's "all quite simple". "Simple rules produce complex behaviour". "Poof".
Well, yeah, if eleven billion years is "poof" to you. Or are the premises in a-life research irrelevant to physical science? Am I missing some nuance because I'm not in the field?
John again:
You can't get something from nothing.
You can't blow shit up and get DNA. It doesn't work that way.
Big Bang -> us. No?
John once more:
The Universe is god and is clearly responsible for everything that is, has been, and ever will be.
Waitwaitwait. Can you clarify your use of "god" there? Do you mean simply that the Universe is self-originated and self-governing? If you apply the g-word in that case, you seem to be advocating scientific pantheism -- which is a very nice philosophy, but isn't atheism.
I am interested in what you're saying, John. I just want to make sure I understand your terms.
Some replies on this page have been deleted or are under review.
P.S.
It took some six years but artificial life has been created. Craig Venter unveiled synthetic life in 2010. He had a precursor talk at ted.com. Wikipedia has an informative entry on the subject. Also exists an entry specifically on Craig's field. The public is only slowly waking up to the achievement.
No comments:
Post a Comment