Monkey Men vs. Bible Thumpers (Evolution vs. Creationism)

[mico saudad]

Wait, you've got me confusified. What is "irreducible complexity?" Why is it interesting and what does its existence prove?

What side of things is Intelligent Design on and what are they trying to prove by their little thought experiments in computer science? I can see arguments being made either way. If they want to stir up controversy around algorithms, then why aren't they all up with undecidability? That's way more popular than search for this sort of thing.

Irreducible complexity. Have you ever played the game Jenga? First you build up a stack of blocks and then after it's built you take turns pulling one out until the leaning tower falls over. If you look at a newly constructed Jenga tower, it's pretty easy to see how it was made. You can have a tower that is one block tall or ten feet tall, but it all is easily reduced to the single block and some basic rules of construction.

This is reducible complexity.

irreducible complexity is where you have a series of interconnected systems that each rely on the others in order to function correctly. Imagine an airplane without wings or without landing gear or without a feul tank. An airplane is in many ways irreducibly complex*.

The reason the concept of irreducible complexity is so important is that it is the last major hurdle for the theory of evolution. If a thing is irreducibly complex, then how is possible to get there by unguided natural selection?

Creationists seize on this challenge as a means of discrediting evolution. They say that the current inability of scientists to explain how irreducible complexity emerges from natural selection constitutes a failure of the theory to describe nature, and as such other theories that describe the origin of humans should also be taught in our schools.

This is why the concept of irreducible complexity is so important.

My point was that recent work with computer models shows that irreducibly complex systems are a natural byproduct of evolution. Therefore the theory of evolution by natural selection is able to explain irreducible complexity.
______
* edit: But also in many ways it is not irreducibly complex. Remove the feul tank and engine and it could be a glider. Remove the wings and it could be some means of ground transportation. The same is true for all life. If you take away your hair and you have a <a href="http://www.sptimes.com/2005/02/04/World ... l">genetic abnormality that fuses your feet together</a> then you might be pretty good when the world floods

[erik]

1. Human creates program that allows population to exist and individuals to exhibit variation and reproduce based on their fitness. Call this program 'Variation'.
2. Human writes a program that rewards 'circuit-like' behaviour and punishes non-circuit-like behaviour. Special bonuses are given to individuals that come close to being able to calculate square roots. The degree of reward or punishment corelates exactly with the amount of reproductive success the individual will have. Call this program 'Selection'.
3. Human writes a third program that will run as follows:
- run 'Variation'
- run 'Selection'
- repeat 10,000 times
- display population of the 10,000th generation
4. Human runs the program and comes back later to see what the 10,000th generation looks like.

Someone intelligent designed an experiment that favored square root calculators, and ended up with alot of square root calculators. Would the population be filled with square root calculators if there had not been some intelligent person who designed these programs to do so? This seems to me to do nothing to disprove the intelligent design people at all.

[Mogosagatai]

Someone intelligent designed an experiment that favored square root calculators, and ended up with alot of square root calculators. Would the population be filled with square root calculators if there had not been some intelligent person who designed these programs to do so? This seems to me to do nothing to disprove the intelligent design people at all.

The intelligent design people believe, specifically, that there was a guiding hand throughout the process of evolution, rather than an unguided process that spawned from some initial setup. That experiment does nothing to disprove an initial creator, and abecedarian has admitted as much, but it does give a heap of evidence that evolution can occur without a guiding hand.

One can still easily argue that God simply set the whole thing up and then let the program (natural selection) run, knowing the output it would produce (the world as we know it today).

[mico saudad]

Someone intelligent designed an experiment that favored square root calculators, and ended up with alot of square root calculators. Would the population be filled with square root calculators if there had not been some intelligent person who designed these programs to do so?

Well it would take an intelligence to determine that the only suitable answer is <x> specific answer.

There are other scientists that allow 'in silico' evolution to progress without specifying an end product. In each of these cases complexity develops perfectly well without an intelligence specifying an outcome. All it takes is the system to be set up.

And no one can say jack (scientifically) about how our system came to be set up the way it was. So my point still stands that there is no observational difference between the scientific theory of evolution's description of nature and one that says that god set the theory of evolution into motion.

So like Mogosagatai says, without an observational difference, anyone using the scientific method must beg out of addressing that question.

We're only interested in accurate descriptions and predictions. And ID is neither a valid description according to current data, nor does it have any predictive utility that the theory of evolution lacks.

[erik]

The intelligent design people believe, specifically, that there was a guiding hand throughout the process of evolution, rather than an unguided process that spawned from some initial setup. That experiment does nothing to disprove an initial creator, and abecedarian has admitted as much, but it does give a heap of evidence that evolution can occur without a guiding hand.

One can still easily argue that God simply set the whole thing up and then let the program (natural selection) run, knowing the output it would produce (the world as we know it today).

I'm not an Intelligent Designer, and I think they're dumb and unscientific. But I can't find anything that says that all IDers believe that there was a guiding hand over a long period of time to produce complex organisms. From what I can tell they seem to believe that "certain features of the universe and of living things are best explained by an intelligent cause rather than an undirected process such as natural selection. ID is thus a scientific disagreement with the core claim of evolutionary theory that the apparent design of living systems is an illusion."

But this experiment is not an example of natural selection. It's an artificial selection. The dude running the program has done things to promote a characteristic that he wants to promote: square root computation. He is God picking out the thing that he wants to show up. It's not a bunch of mutations deciding for themselves which mutations are going to be the ones that should show up more.

They are still dumb and unscientific.

[bridge]

So do ID theorists believe that the complexity of the body and each of the (irreducibly complex?) systems/processes/etc that go on in it are all individually too complex to have evolved without guidance? Or that someone/thing must have helped particular systems develop simultaneously because they are so interdependent?

There are obviously theories as to how some (most? all? I have no idea how many) of the complex systems in the body developed. Do ID people agree with these theories but think they couldn't have happened on their own, or just totally discount the theories?

Take the respiratory system and evolution of lungs from gills. The theory I've heard explaining this is that some primitive fish in O2 poor stagnant waters gulped air at the surface of water, which went into their stomachs and was absorbed into the circulatory system. Fish that thought to do this were selected for and eventually an outpocketing of the stomach/new branch of the circulatory system to this new part of the body eveolved, and from this lungs evolved, a circulatory system that began to bypass the gills and a two (fish), then three (amphibians), then four (mammals...) chambered heart, etc.

It seems sort of ridiculous to think that fish gulping air at the surface of water led to land-dwelling animals... But having a hard time grasping a concept doesn't mean the process and processes like it couldn't have occured on their own. Obviously there were thousands/millions/billions of yrs, which is a span of time I definitely can't wrap my brain around, for all of these things to occur. What determines if something is too complex to have occured on its own?

[mico saudad]

Erik here's how I would argue against ID if I were you:

If you believe the computer data is evidence against ID then there's your argument, but even if you don't buy the computer simulations, then that means that their hypothesis that certain things are 'too complex' is untestable. If it is not debunkable then it is not science.

gills/lungs,... and what determines if something is too complex to evolve on its own

Well you chose an interesting example, because there are drought-plagued areas. At certain times of the year the bodies of water that they live in dry out. <a href="http://www.colszoo.org/animalareas/shor ... l">Certain fish known as lungfish</a> have the ability to survive for months in a sort of hibernation without any water whatsoever.

When you think about having to survive your puddle of water drying up it makes sense that the ability to survive in air would be useful. In fact many kinds of fish have found many ways to survive on land. Mudskippers carry water with them while they live on land and reoxygenate it while other amphibious fish can pass air through their gills. DNA evidence shows that the ancestors of the lungfish (and not these other fish) are likely our last common ancestor with frogs and salamanders.

And you ask a great question. There is no scientific model that exists that can possible measure the degree of dificulty that it would take evolution to generate a given structure. You're absolutely right that much of the so-called complexity that ID-propoenents hold up is merely due to a lack of a scientific model to gauge complexity.

[Adam!]

Didn't the programmer select the circuits, albeit blindly? It's not like they put a motherboard down on a table, came back a week later and it was a square root generator, is it?

I thought that humans were selecting circuits that came close to approximating a square root function. That's intelligent interference.

Someone intelligent designed an experiment that favored square root calculators, and ended up with a lot of square root calculators. Would the population be filled with square root calculators if there had not been some intelligent person who designed these programs to do so? This seems to me to do nothing to disprove the intelligent design people at all.

But his experiment is not an example of natural selection. It's an artificial selection. The dude running the program has done things to promote a characteristic that he wants to promote: square root computation. He is God picking out the thing that he wants to show up. It's not a bunch of mutations deciding for themselves which mutations are going to be the ones that should show up more.

Took a class on this. To explain the difference our prof used the example of evolving Scheme applications. Scheme [corrected. Thanks bzl] is a programming language that (due to its syntax) is guaranteed to compile, even if it is randomly generated gobbledygook. At every generation all of the functions (and all of the e. coli) are evaluated using a fitness function: the higher the score on the fitness function the more offspring they add to the next generation.

Natural Evolution of e. coli:
Initial Population: A Petri-dish full of bacteria
Variance: Random mutation
Fitness Function: The better a bacterium is at reproduction the more fit it is.
End Result after 10,000 generations: A Petri-dish full of bacteria that are [ideally] better equipped to reproduce

Simulated Evolution of a Scheme Algorithm:
Initial Population: 10,000 randomly generated compiling Scheme functions
Variance: Random mutation
Fitness Function: The closer a function's regression is to the square-root function the more fit it is.
End Result after 10,000 generations: A bunch of Scheme square root functions.

The two concepts are identical, but people often get hung up on the fitness function. In nature living thingies (and some non-living thingies) have the built in fitness function of reproduction. Because Scheme programs do not fuck, a fitness function must be defined. A programmer does write this, but if they did not then simulated evolution would be nothing like natural evolution because it would have no way of differentiating how well adapted different specimens are.

[mico saudad]

description

Good description puce

[erik]

The two concepts are identical, but people often get hung up on the fitness function. In nature living thingies (and some non-living thingies) have the built in fitness function of reproduction. Because Lisp programs do not fuck, a fitness function must be defined. A programmer does write this, but if they did not then simulated evolution would be nothing like natural evolution because it would have no way of differentiating how well adapted different specimens are.

The two concepts are similar, they aren't identical. If one says "My evolution simulation requires that an intelligent being write programs and define what fitness is, which therefore provides evidence that natural evolution doesn't require an intelligent designer", then the two things are far enough apart to not have one provide evidence for the other.

[Adam!]

The two concepts are identical, but people often get hung up on the fitness function. In nature living thingies (and some non-living thingies) have the built in fitness function of reproduction. Because Lisp programs do not fuck, a fitness function must be defined. A programmer does write this, but if they did not then simulated evolution would be nothing like natural evolution because it would have no way of differentiating how well adapted different specimens are.

The two concepts are similar, they aren't identical. If one says "My evolution simulation requires that an intelligent being write programs and define what fitness is, which therefore provides evidence that natural evolution doesn't require an intelligent designer", then the two things are far enough apart to not have one provide evidence for the other.

If you like you could theoretically code applications that simulate the behavior of e. coli bacteria, then make the fitness function be how well a virtual bacterium reproduces. With enough detail you could make something observationally identical to the same experiment conducted with actual bacteria. This would be really no different than the square root example.

Intelligent Design necessitates that irreducibly complex structures cannot arise in an evolutionary system that involves an initial population, a variance and a fitness function, regardless of the nature of these three components. Examples of evolutionary algorithms have been coded (hell, I've coded one) that contradict this claim.

The fact that a computer and a programmer are involved is neat, but it doesn’t invalidate this counterexample.

[bz£]

I should probably quit repeating myself (and set an example!) but, again, Puce's example is just another form of search, which is one of the most basic things you might want to write a computer program to do. (For what it's worth, basically every problem in AI can be expressed as a search problem.)

The lisp stuff is inaccurate, which hurts me to the quick, given all the years I spent writing lisp code for a salary. It doesn't matter, really, though, the analogy still works. You could use, say, java, instead, and the argument doesn't change, except that most smart programmers would refuse to write this sort of thing in java because it would take way too much effort.

I guess you could argue that nature provides an (indirect) definition of fitness: the strong survive, the adaptable survive, the rest don't. Maybe that comes down from god and maybe it's just a consequence of the way things are; I s'ppose we'll never know.

[Adam!]

I haven't burnt my notes from that class yet (next weekend, baby!) so I checked them out and you're right: I was thinking of Scheme, a subset of Lisp with ultra-simple semantics. However, I still believe you can use a grammar to randomly generate semantically correct gobbledygook Lisp programs. In fact, I think I have the code lying around somewhere.

And yes, you're very right that evolutionary programming is still just dumb hill climbing. But it's a way to traverse a search space that sounds so nifty that getting you're lab a government grant is a synch.

[erik]

Intelligent Design necessitates that irreducibly complex structures cannot arise in an evolutionary system that involves an initial population, a variance and a fitness function, regardless of the nature of these three components. Examples of evolutionary algorithms have been coded (hell, I've coded one) that contradict this claim.

Is a square root function irreducibly complex? What does the phrase "irreducibly complex" (which in general has a nebulous meaning) refer to when discussing math algorithms?

[Adam!]

Intelligent Design necessitates that irreducibly complex structures cannot arise in an evolutionary system that involves an initial population, a variance and a fitness function, regardless of the nature of these three components. Examples of evolutionary algorithms have been coded (hell, I've coded one) that contradict this claim.

Is a square root function irreducibly complex? What does the phrase "irreducibly complex" (which in general has a nebulous meaning) refer to when discussing math algorithms?

An easy way to think about irreducible complexity is by thinking about a mousetrap, which is made out of 4 or so parts. Take any of those parts away and you have a broken mousetrap. Something is irreducibly complex if when you reduce it's complexity further it becomes non-functional or something else entirely.

There is a class of algorithms (called NP-Hard) which most likely cannot be reduced to polynomial complexity. If you find a way, run to Stockholm to collect your Nobel Prize.

The simplest algorithm for calculating a square root is a recursive function that uses two divides and a sum. Take any part of that away and the algorithm diverges wildly from the square root function. It's a tough call, but I'd say that counts as irreducibly complex. There is a formula for tabulating the minimum complexity of an algorithm, but I'd rather pull my eyes out then try to remember it.

For a fun time, check out this.

[jack]

proof of god

[mico saudad]

Some kid in my neighborhood was doinga school project on this debate and was walking around asking people to go on camera with their opinions during a yard sale I was having. One old guy with a little black Schnauzer started talking into the camera and said:

"Evolution versus Creationism? Well one is bullshit, the other is fact."

I thought it was an artfully ambiguous answer.

[roymond]

Good article from the New Yorker on the Intelligent Design movement

I'm always curious: if God put Us here for His little experiment. What's all the rest of the universe for? Is it there for Us to ponder and therefore become overwhelmed and more easily turn to God as the answer?

[jack]

Good article from the New Yorker on the Intelligent Design movement

I'm always curious: if God put Us here for His little experiment. What's all the rest of the universe for? Is it there for Us to ponder and therefore become overwhelmed and more easily turn to God as the answer?

of course God knows the answer, being all omnipotent and all.

if we had all the answers, there would be no questions. and life would be boring without questions.

[thehipcola]

Some kid in my neighborhood was doinga school project [snnnnnip]

are you sure he wasn't Dinga DaDonga the school project? HAHAHAHHHAHa...haahahhaahhha....haa hhah...ha....

ha...

...ahem..

[looks around, walks away quietly.....]

[erik]

if God put Us here for His little experiment. What's all the rest of the universe for?

Control group.

[jack]

for a minute there rob, i thought you were tom.