Total Posts:21|Showing Posts:1-21
Jump to topic:

Life Game

slo1
Posts: 4,364
Add as Friend
Challenge to a Debate
Send a Message
10/29/2014 7:45:21 AM
Posted: 2 years ago
This is cool.

I'm amazed that there is enough predictability that it would be possible to build a functional computer within the game of life.

It goes to show that we often mis-diagnose "random" and carelessly use the word in many instances when things are not really random.

When there are natural rules that things follow stuff can get organized in weird ways.
v3nesl
Posts: 4,505
Add as Friend
Challenge to a Debate
Send a Message
10/29/2014 10:16:36 AM
Posted: 2 years ago
At 10/29/2014 7:45:21 AM, slo1 wrote:
This is cool.

I'm amazed that there is enough predictability that it would be possible to build a functional computer within the game of life.

It goes to show that we often mis-diagnose "random" and carelessly use the word in many instances when things are not really random.

When there are natural rules that things follow stuff can get organized in weird ways.

Yeah, this line is a bit misleading: In most cases, it is impossible to look at a starting position (or pattern) and see what will happen in the future. In the mathematical sense, it IS possible to see everything that will happen from a given starting position. There's nothing random about it at all.
This space for rent.
dylancatlow
Posts: 12,255
Add as Friend
Challenge to a Debate
Send a Message
10/29/2014 11:19:39 AM
Posted: 2 years ago
At 10/29/2014 7:45:21 AM, slo1 wrote:
This is cool.

I'm amazed that there is enough predictability that it would be possible to build a functional computer within the game of life.

It goes to show that we often mis-diagnose "random" and carelessly use the word in many instances when things are not really random.

When there are natural rules that things follow stuff can get organized in weird ways.

It's hypnotizing to watch. If you haven't already, check out his collection of life forms. Some of them are amazing.
Enji
Posts: 1,022
Add as Friend
Challenge to a Debate
Send a Message
10/29/2014 11:55:57 AM
Posted: 2 years ago
At 10/29/2014 10:16:36 AM, v3nesl wrote:
Yeah, this line is a bit misleading: In most cases, it is impossible to look at a starting position (or pattern) and see what will happen in the future. In the mathematical sense, it IS possible to see everything that will happen from a given starting position. There's nothing random about it at all.

Actually, it's correct; in the mathematical sense it is NOT possible to predict the outcome from any given starting position. This is because Conway's game of life has been proven to be Turing-complete, and computability theory proves that it is impossible to determine what a Turing-complete program will do in the general case over an arbitrarily long time. For example, there's no algorithm which can correctly tell you whether a program will eventually halt or continue forever in every case; there's no algorithmic short-cut to tell you the eventual outcome for any given starting arrangement. As they put it, "The only way to find out is to follow the rules of the game."
v3nesl
Posts: 4,505
Add as Friend
Challenge to a Debate
Send a Message
10/29/2014 1:42:07 PM
Posted: 2 years ago
At 10/29/2014 11:55:57 AM, Enji wrote:
At 10/29/2014 10:16:36 AM, v3nesl wrote:
Yeah, this line is a bit misleading: In most cases, it is impossible to look at a starting position (or pattern) and see what will happen in the future. In the mathematical sense, it IS possible to see everything that will happen from a given starting position. There's nothing random about it at all.

Actually, it's correct; in the mathematical sense it is NOT possible to predict the outcome from any given starting position. This is because Conway's game of life has been proven to be Turing-complete, and computability theory proves that it is impossible to determine what a Turing-complete program will do in the general case over an arbitrarily long time. For example, there's no algorithm which can correctly tell you whether a program will eventually halt or continue forever in every case; there's no algorithmic short-cut to tell you the eventual outcome for any given starting arrangement. As they put it, "The only way to find out is to follow the rules of the game."

okey dokey. Never mind, lol.
This space for rent.
dylancatlow
Posts: 12,255
Add as Friend
Challenge to a Debate
Send a Message
10/29/2014 4:38:19 PM
Posted: 2 years ago
I think there are a number of problems with comparing this to abiogenesis. In the life game, you don't need a complex starting configuration to get something going, while in actual life the starting configuration required is as complex as life itself (comparing a contrived dot pattern to a living organism is kind of a joke). There's also the fact that in the life game, interactions between cells are limited to just a few options, while in real life matter and energy interact with far more complexity. And of course, the organism's environment is basically non-existent. They don't have to find food or configure themselves/adjust their behavior to external conditions or have any sort of organs. Nor do the life forms have to reproduce themselves, nor are there any choices when it comes to what the organisms are composed of (there's just one kind of material which behaves in exactly the same way as all the other material), nor is something like DNA involved. The conditions under which life would have had to transpire are so different that there's just no reason to compare them. It's like trying to probe the question "does communism work" with a party game.
RainbowDash52
Posts: 294
Add as Friend
Challenge to a Debate
Send a Message
10/29/2014 8:24:56 PM
Posted: 2 years ago
At 10/29/2014 4:38:19 PM, dylancatlow wrote:
I think there are a number of problems with comparing this to abiogenesis. In the life game, you don't need a complex starting configuration to get something going, while in actual life the starting configuration required is as complex as life itself

Actually, since Conway's game of life is turing complete and the laws of physics can be simulated by a turing machine, then if the laws of physics allows for starting off with a simple configuration and resulting in complex configuration, then the same can be true for Conway's game of life.

(comparing a contrived dot pattern to a living organism is kind of a joke). There's also the fact that in the life game, interactions between cells are limited to just a few options, while in real life matter and energy interact with far more complexity. And of course, the organism's environment is basically non-existent. They don't have to find food or configure themselves/adjust their behavior to external conditions or have any sort of organs. Nor do the life forms have to reproduce themselves, nor are there any choices when it comes to what the organisms are composed of (there's just one kind of material which behaves in exactly the same way as all the other material), nor is something like DNA involved. The conditions under which life would have had to transpire are so different that there's just no reason to compare them. It's like trying to probe the question "does communism work" with a party game.

Conway's game of life is an overly simplified comparison of life, but that doesn't mean the comparison is not useful. Real life is extremely complex and we have a very limited understanding of it. So having a simplified version of something complex can have some educational value, even if it doesn't perfectly answer everything.
RoyLatham
Posts: 4,488
Add as Friend
Challenge to a Debate
Send a Message
10/29/2014 10:58:52 PM
Posted: 2 years ago
I think the game is at least suggestive of mechanisms of abiogenesis. The thing to notice is that something does not need to be complex to reproduce and evolve. Any molecule that draws energy from the environment to reproduce itself is subject to evolution. A mutated version that uses energy more efficiently, reproduces faster, or is better able to survive the environment will dominate over less capable molecules. Viruses are borderline cases between molecules and living things.

One of the theories of abiogenesis involves a natural crystal face providing molecular sites to organize naturally occurring organic radicals. That's kind of like the life game, where something emerges from random placement.
v3nesl
Posts: 4,505
Add as Friend
Challenge to a Debate
Send a Message
10/30/2014 8:04:09 AM
Posted: 2 years ago
At 10/29/2014 8:24:56 PM, RainbowDash52 wrote:
At 10/29/2014 4:38:19 PM, dylancatlow wrote:
I think there are a number of problems with comparing this to abiogenesis. In the life game, you don't need a complex starting configuration to get something going, while in actual life the starting configuration required is as complex as life itself

Actually, since Conway's game of life is turing complete and the laws of physics can be simulated by a turing machine, then if the laws of physics allows for starting off with a simple configuration and resulting in complex configuration, then the same can be true for Conway's game of life.

Who says it's complex? We see interesting patterns in the dots, but that may be a reflection of our complex brains, not the complexity of the dots. The algorithm is what it is - to correctly measure the complexity of the game you must measure the algorithm, not the dots.

Read closely and you'll see that Conway spent time and testing to derive an algorithm that produced stuff we find interesting. That's analogous to intelligent design and genetics, I'd say. And we could add some interest to the game, I bet, by adding a random, i.e. noise component - a roll of the dice somewhere. I bet we would model 'statistically bounded variation', a bell curve kind of thing, if you measured say, run-length-before-repeating or maximum-pattern-dots or something like that.

I think there is this desire to find magic in order to justify the belief in magic that is Darwinian evolution. But this ain't it. No magic here, just something interesting.
This space for rent.
dylancatlow
Posts: 12,255
Add as Friend
Challenge to a Debate
Send a Message
10/30/2014 10:35:06 AM
Posted: 2 years ago
At 10/29/2014 10:58:52 PM, RoyLatham wrote:
I think the game is at least suggestive of mechanisms of abiogenesis. The thing to notice is that something does not need to be complex to reproduce and evolve. Any molecule that draws energy from the environment to reproduce itself is subject to evolution. A mutated version that uses energy more efficiently, reproduces faster, or is better able to survive the environment will dominate over less capable molecules. Viruses are borderline cases between molecules and living things.

One of the theories of abiogenesis involves a natural crystal face providing molecular sites to organize naturally occurring organic radicals. That's kind of like the life game, where something emerges from random placement.

I'd agree that it's suggestive, but only in a superficial sense. In the actual world, reproduction and evolution are astronomically more complex. The controversy hinges on irreducible complexity, and I don't think the life game is sufficiently similar to be of scientific relevance.
RoyLatham
Posts: 4,488
Add as Friend
Challenge to a Debate
Send a Message
10/30/2014 12:02:34 PM
Posted: 2 years ago
At 10/30/2014 10:35:06 AM, dylancatlow wrote:
I'd agree that it's suggestive, but only in a superficial sense. In the actual world, reproduction and evolution are astronomically more complex. The controversy hinges on irreducible complexity, and I don't think the life game is sufficiently similar to be of scientific relevance.

So you think that any molecule that replicates itself must be too complex to occur in nature? That's a religious belief, not a scientific one. I don't see why it couldn't be quite simple. None of the other attributes of life are required.
dylancatlow
Posts: 12,255
Add as Friend
Challenge to a Debate
Send a Message
10/30/2014 12:05:24 PM
Posted: 2 years ago
At 10/30/2014 12:02:34 PM, RoyLatham wrote:
At 10/30/2014 10:35:06 AM, dylancatlow wrote:
I'd agree that it's suggestive, but only in a superficial sense. In the actual world, reproduction and evolution are astronomically more complex. The controversy hinges on irreducible complexity, and I don't think the life game is sufficiently similar to be of scientific relevance.

So you think that any molecule that replicates itself must be too complex to occur in nature? That's a religious belief, not a scientific one. I don't see why it couldn't be quite simple. None of the other attributes of life are required.

I didn't say that. I said that the conditions of the life game are not comparable to the conditions of the real world, and thus the life game is not an appropriate tool for simulation. It's simply irrelevant.
RoyLatham
Posts: 4,488
Add as Friend
Challenge to a Debate
Send a Message
10/30/2014 12:27:18 PM
Posted: 2 years ago
At 10/30/2014 12:05:24 PM, dylancatlow wrote:
I didn't say that. I said that the conditions of the life game are not comparable to the conditions of the real world, and thus the life game is not an appropriate tool for simulation. It's simply irrelevant.

One point of relevance is that it defies the claim that something complex cannot originate on its own from something simple.

No simulation ever captures the full complexity of the real world, but that means that results have to be interpreted carefully, not that they are irrelevant. Life doesn't prove abiogenesis, but it teaches how complexity can arise from very simple conditions.
dylancatlow
Posts: 12,255
Add as Friend
Challenge to a Debate
Send a Message
10/30/2014 1:05:54 PM
Posted: 2 years ago
At 10/30/2014 12:27:18 PM, RoyLatham wrote:
At 10/30/2014 12:05:24 PM, dylancatlow wrote:
I didn't say that. I said that the conditions of the life game are not comparable to the conditions of the real world, and thus the life game is not an appropriate tool for simulation. It's simply irrelevant.

One point of relevance is that it defies the claim that something complex cannot originate on its own from something simple.


That claim was never going to hold up anyway.

No simulation ever captures the full complexity of the real world, but that means that results have to be interpreted carefully, not that they are irrelevant. Life doesn't prove abiogenesis, but it teaches how complexity can arise from very simple conditions.
dylancatlow
Posts: 12,255
Add as Friend
Challenge to a Debate
Send a Message
10/30/2014 1:36:49 PM
Posted: 2 years ago
At 10/30/2014 12:27:18 PM, RoyLatham wrote:
At 10/30/2014 12:05:24 PM, dylancatlow wrote:
I didn't say that. I said that the conditions of the life game are not comparable to the conditions of the real world, and thus the life game is not an appropriate tool for simulation. It's simply irrelevant.

One point of relevance is that it defies the claim that something complex cannot originate on its own from something simple.

No simulation ever captures the full complexity of the real world, but that means that results have to be interpreted carefully, not that they are irrelevant. Life doesn't prove abiogenesis, but it teaches how complexity can arise from very simple conditions.

I think the fact that one can create a recognizable pattern using only 5 cells shows how unlike the situations are. All the life forms are predicated on the condition that cells reproduce and die based on surrounding cells. In the actual world, individual particles are always moving, and have practically no influence on other particles for more than a split second.
v3nesl
Posts: 4,505
Add as Friend
Challenge to a Debate
Send a Message
10/30/2014 1:50:52 PM
Posted: 2 years ago
At 10/30/2014 1:05:54 PM, dylancatlow wrote:
At 10/30/2014 12:27:18 PM, RoyLatham wrote:
At 10/30/2014 12:05:24 PM, dylancatlow wrote:
I didn't say that. I said that the conditions of the life game are not comparable to the conditions of the real world, and thus the life game is not an appropriate tool for simulation. It's simply irrelevant.

One point of relevance is that it defies the claim that something complex cannot originate on its own from something simple.


How does it defy the claim? You'll have to define a measurement for complexity, and since the game was designed by a highly complex human brain, I can't imagine how you're going to define it so that the game of life generates complexity.

I think the math that would appropriate here is information theory, specifically, Shannon's concept of 'entropy', measuring the information content of data. And by that metric, the game of life adds precisely zero entropy. Since all outcomes are a deterministic result of the logic of the game, all outcomes are inherent in the original configuration + algorithm. The complexity exists in the original configuration + algorithm, it is merely expressed by running the game. To think of it another way, the game of life is a lossless data compression algorithm for the pattern of dots produced from any given starting point. You can express a series of dot patterns in a compact form.
This space for rent.
dylancatlow
Posts: 12,255
Add as Friend
Challenge to a Debate
Send a Message
10/30/2014 2:10:09 PM
Posted: 2 years ago
At 10/30/2014 12:27:18 PM, RoyLatham wrote:
At 10/30/2014 12:05:24 PM, dylancatlow wrote:
I didn't say that. I said that the conditions of the life game are not comparable to the conditions of the real world, and thus the life game is not an appropriate tool for simulation. It's simply irrelevant.

One point of relevance is that it defies the claim that something complex cannot originate on its own from something simple.

No simulation ever captures the full complexity of the real world, but that means that results have to be interpreted carefully, not that they are irrelevant. Life doesn't prove abiogenesis, but it teaches how complexity can arise from very simple conditions.

I also think there's a sense in which the life game is going after a strawman. No one is claiming that given the right starting configuration, complexity cannot unfold according to simple rules (at least among those who accept evolution). The problem is that the forms which display self-reinforcing complexity have only been created using intelligent intervention. To my knowledge, nothing impressive has arisen out of "cell soup". Simply because one can configure a complex starting state in paused time does not mean that such a state could arise on its own according to simple rules.
slo1
Posts: 4,364
Add as Friend
Challenge to a Debate
Send a Message
10/30/2014 7:52:50 PM
Posted: 2 years ago
At 10/30/2014 2:10:09 PM, dylancatlow wrote:
At 10/30/2014 12:27:18 PM, RoyLatham wrote:
At 10/30/2014 12:05:24 PM, dylancatlow wrote:
I didn't say that. I said that the conditions of the life game are not comparable to the conditions of the real world, and thus the life game is not an appropriate tool for simulation. It's simply irrelevant.

One point of relevance is that it defies the claim that something complex cannot originate on its own from something simple.

No simulation ever captures the full complexity of the real world, but that means that results have to be interpreted carefully, not that they are irrelevant. Life doesn't prove abiogenesis, but it teaches how complexity can arise from very simple conditions.

I also think there's a sense in which the life game is going after a strawman. No one is claiming that given the right starting configuration, complexity cannot unfold according to simple rules (at least among those who accept evolution). The problem is that the forms which display self-reinforcing complexity have only been created using intelligent intervention. To my knowledge, nothing impressive has arisen out of "cell soup". Simply because one can configure a complex starting state in paused time does not mean that such a state could arise on its own according to simple rules.

True, but think what might happen when the rules change in mid stream due to environmental factors. Next thing you know the game of life is spelling words out at random times.
RoyLatham
Posts: 4,488
Add as Friend
Challenge to a Debate
Send a Message
10/30/2014 8:21:51 PM
Posted: 2 years ago
At 10/30/2014 2:10:09 PM, dylancatlow wrote:
I also think there's a sense in which the life game is going after a strawman. No one is claiming that given the right starting configuration, complexity cannot unfold according to simple rules (at least among those who accept evolution).

This is exactly what creationists claim. Creationists say it impossible for God to have set up the laws of nature and from that alone life evolved. Creationists insist that God had to intervene to make the thing work. That's because God is not smart enough to figure out how to add information to DNA within the bounds of the laws of nature, so divine intervention is absolutely required. The position of non-creationist believers is that God figured out how to work through evolution.

The problem is that the forms which display self-reinforcing complexity have only been created using intelligent intervention. To my knowledge, nothing impressive has arisen out of "cell soup". Simply because one can configure a complex starting state in paused time does not mean that such a state could arise on its own according to simple rules.

One of the life game questions was whether a random starting pattern could grow forever. Experimenters discovered that a random starting pattern could produce a glider generator, a pattern that generates a stream of gliders that move off the screen and continue forever. I guess you can argue whether that is impressive or not. I think it is at least surprising.
v3nesl
Posts: 4,505
Add as Friend
Challenge to a Debate
Send a Message
10/31/2014 4:29:49 PM
Posted: 2 years ago
At 10/30/2014 8:21:51 PM, RoyLatham wrote:
...

One of the life game questions was whether a random starting pattern could grow forever. Experimenters discovered that a random starting pattern could produce a glider generator, a pattern that generates a stream of gliders that move off the screen and continue forever. I guess you can argue whether that is impressive or not. I think it is at least surprising.

Yeah, you're really not thinking about this correctly, and I guess you wouldn't listen to me anyway, I'm a creationist, so it really doesn't matter that computing machines are my thing.

So making dots that continue forever is no big deal. The simple algorithm "add one dot to right of the line" does that. Perhaps the hangup in your brain is that you don't want to face that if there were indeed some natural algorithm we could call "evolution", we'd be able to run it. We could run it at will, we could evolve at will. But, of course, no such algorithm has been shown to be possible.
This space for rent.