What is a GA?
Discuss.
704 thoughts on “Holding tank for general chatter about GAs”
Leave a Reply
You must be logged in to post a comment.
The Skeptical Zone
"I beseech you, in the bowels of Christ, think it possible that you may be mistaken."
What is a GA?
Discuss.
You must be logged in to post a comment.
Design ‘theory’ in a nutshell.
So if GAs are so powerful and mimic natural selection, it seems to me that you are arguing that evolution can in fact do everything claimed of it, no ID required.
Where does ID come into this scenario Joe?
Well cars behave the way they do because that is how they were designed- and as dawkins and others have said if the organisms are designed then we are looking at a totally different kind of biology
Presumably then that’s also true for HIV and other similar organisms that cause untold misery to the human race?
Mice stupid response- I said GAs TRY to mimic NS- they don’t because NS doesn’t do anything
No, It was John Paul, or his Dad. DONT MENTION HIS DAD!
Only in your little evotard mind did I do any flopping.
IOW tard you are a moron.
How odd! A search heuristic, based on a natural process that doesn’t work, does work.
It is supposed to be based on natural processes but it isn’t as natural processes do not have a goal.
As I said you are a moron and you think your ignorance means something.
What’s the point of mimicking something that does nothing.
GAs solve problems. NS solves problems.
If GAs are copied from nature, and demonstrably work, the more you argue for their power the more you argue that NS can in fact do what is claimed for it.
Best.Thread.Ever.2.
Survival is not a goal? How so?
Let’s have us some Wikipedia, Joe:
If you read beyond the first sentence, you find out that GAs are not used for the purpose of simulating biological evolution. Their purpose is to “generate useful solutions to optimization problems.” They are “inspired” by the principles of Darwinian evolution.
Cars don’t reproduce.
It’s funny, but in the past you’ve claimed that things like HIV exist because the “design got corrupted”, an allusion to “the fall”.
Yet these cellular GAs don’t appear to suffer from that corruption. They are working exactly as originally designed.
Seems to me if the designer can do that it can also keep HIV from becoming a problem.
Funniest.
Reply.
Evah.
So when AVIDA is used as evidence that complex biological functionality can evolve- Tom Schneider is lying?
That is what I said too. Thanks.
Reproduction is something your position can’t explain.
Joe,
So if GAs “direct” mutations by natural selection (fitter variants survive preferentially) they why cannot evolution do the same?
How is it a goal? Make your case.
GAs don’t use natural selection- you are an ignorant wanker.
Shubin found Tiktaalic but was wrong because he found found the right thing in the wrong place, even though it was there!
GAs (which live in the cell) model evolution and although they work, evolution doesn’t. But I’m not anti-evolution!
Rabid assclown. Mash that keyboard, culture warrior.
NS doesn’t solve anything- if NS did something you would have evidence to support that claim- but you don’t.
Let’s call the number of things my position explains X.
Let’s call the number of things that ID explains Y.
X>0
Y=0
Therefore X is more useful then Y.
How does *your* position explain reproduction?
“Reproduction is how things were designed to be”.
And more RichTARD ignorance- intelligent design evolution works- blind watchmaker evolution doesn’t.
Ooh. Goal is riddled with intentionality!!!!!1111one eleveny. But it is what evolution DOES.
Hey asshole- IF YOUR position had some positive evidence there wouldn’t be any ID
Oooooh! TE, now, YEC?
Rich- you couldn’t support the claims of your position if your life depended on it.
What does the selection in a GA Joe?
There *is* no ID Joe. Just some people writing a few blogs and books. No actual science.
It does. Read the manual, Joe. Here is John Holland, who is regarded as the father of GAs:
Goal/consequence. Same difference.
Well there doesn’t seem to be a “theory” of evolution- no science at all.
Joe,
Is your GA in every cell or just one?
That would be the GA that is completely contained within the cell, has no knowledge of the fitness of any other cells or animals in the population, and has no way to control the external selection pressure.
How’s that work again Joe?
Ever thought about explaining *why* somebody is wrong rather then just telling them and insulting them?
Joe,
Over at UD Eric says:
Yet
i) You claim that your GA can solve real biological problems, e.g. create new proteins.
ii) You have previously (I asked on this very thread) noted that the solution is not smuggled in by the programmer. The problems really are solved by the GA.
Will you be explaining to Eric why he’s wrong about both those points or not?
You’ve already responded on that thread after that comment, why have you not taken the opportunity to correct Eric on his misunderstandings?
And Joe,
When your GA creates a new protein, does the CSI got up, down or stay the same overall in the cell?
Now, just a minute here – literally!
I put up that post with the link about proto-ribosomes at 1:56 pm local time
You, Joe, replied at 1:57 pm local time
That link was to a full-text paper, not just an abstract. You cannot have read and understood it in a maximum of 2 minutes.
Admit it, Joe – you’re just on automatic here – anything a non-IDist posts, and you object to it and vilify it INSTANTLY, without even giving it a moment’s attention
Way to go, Joe – read nothing, learn nothing.
I’m not usually one for insulting people over the internet: but the word “poltroon” comes to mind.
I think the problem here, and the source of Joe’s frustration, is that his thought processes are so advanced that we normal people just aren’t able to grasp the key concepts. For example, we think of genetic algorithms as things that mimic or are intended to represent natural (i.e., Blind Watchmaker) processes. Because Joe is adamant in his contention that such processes don’t exist, and there is no evidence for their existence, his GAs are based on something else. It’s the “something else” that Joe sees and we don’t that’s causing at least part of the confusion and Joe’s consternation. Maybe he will be good enough to explain that for us, slowly and patiently, so that we can try to keep up.
Another interesting aspect of Joe’s paradigm-shattering, iconoclastic breakthrough is that while he’s been telling us for a long time, when questioned about the mechanism(s) of ID, that design is a mechanism. This is another concept that our comparatively feeble minds have been unable to grasp. It now becomes even more complicated by the fact that Joe seems to be telling us that genetic algorithms are the mechanisms behind ID. So does ID have an actual mechanism beyond design itself? Can you help us with this, JOe?
I hate to sound like a concern troll, but I’ve got to disagree again. This thread is valuable in that it illustrates a core problem with common blog forum software, namely the ease with which potentially productive discussions can be derailed by a single participant (with considerable and enthusiastic assistance from others).
We’ve taken some major steps backward from the capabilities of Usenet newsreaders in the 1990s. If anyone knows of WordPress tools that provide that kind of functionality, please respond on the topic I started in the Penguin Colony.
Yes, of course, Joe. If we knew the solution we wouldn’t have to write a GA.
We write the problem; we do not write the solution. That’s what I’ve been trying to explain.
Your idea is interesting, but it’s not what is normally called a GA – a Genetic Algorithm.
Sigh. I just did. I presented evidence of random processes generating novel proteins.
ALL GAs are based on Darwinian pronciples – sorry, principles. That’s where they got the fucking idea from!
WHAAAT? To refresh the context, you said that producing more offspring (ie NS) does not generate changes in protein. I said mutation does that. And your triumphant answer to that is that, since variation is required for Natural Selection, mutation is part of NS? OK, if it is, NS can generate changes in protein (via its mutation component). If it isn’t, the separate process of mutation is what does the job, not the having-more-offspring part. It’s really not that hard to grasp, you know, this biology thing.
The paper I linked does precisely that – generate novel proteins by ‘blind and undirected processes’. No questions were begged during the course of this research. They looked to see what would happen and reported what did happen. Obviously, the experimenters had to generate the conditions for the experiment, but the rest of it was entirely a demonstration of that very point. Did you even read it?
A simple analogy might help.
A genetic algorithm is akin to a sports competition. The athletes are not asked to achieve a particular result, e.g., run 26 miles in less than 3 hours. The winner simply has to run faster than everybody else.
Contrast this to a job fair where employers select people with specific skills. That’s a predetermined target. Genetic algorithms don’t have one.
Joe, then where did the GA idea come from in the first place then? What was it based on, if not Darwinian pronciples?
What inspired it?
Joe, “fitness function” doesn’t mean the same thing as “selection coefficient”, although they are related. They aren’t interchangeable.
The mutations aren’t “directed” at all, but they are selected if they are beneficial to the phenotype – in other words they will tend to become more prevalent in the population. But that is straight-up Darwinian evolution!
I already said you need to make explicit the biological analogy, otherwise it’s pure assertion. Let me try this… primary growth in plants is governed by thermodynamic cycles of gases. Watt steam engines were designed to make use of such cycles and they do their job just fine. Ergo, I have a testable model for my claim about primary growth. That doesn’t quite work, does it?
OK, I have spent a bit of time moving posts that seem to me to violate this site’s game rules.
It was a long job, and I didn’t get to the end. I also probably moved some stuff by accident, and kept some stuff that should have moved.
But I think that will do for now.
Joe, most of the moved comments were yours, and they were moved because they violated the rules. Please go read them.
It may well be that you felt provoked, but the rules are not waived because of perceived or actual provocation.
The whole point of having this site is to try to establish a site were people of radically opposing views can try to listen to the “other side” and perhaps understand where they are coming from. That means assuming the other person is posting in good faith.
Please can everyone try to sit on their hands when temptation strikes and post in as parliamentary manner as possible.
Cheers
Lizzie
It still appears that the Fundamental Misconception of ID/creationism is still in play with Joe G; namely, it would all be chaos among atoms and molecules if it weren’t for some program guiding them toward some goal.
There is no goal. As for a program or “algorithm” – however Joe G understands the meaning of those words – there aren’t any of those either.
However, there are very, very, very strong electromagnetic interactions among atoms and molecules and the rules of quantum mechanics that constrain the patterns that emerge from these electromagnetic interactions.
Here are some reference points.
Chemical bonds are in the energy range of about 1 electron volt (eV).
The binding energies of solids like metals is on the order of 0.1 eV
The binding energy of liquid water is on the order of 0.01 eV.
There are processes going on in living organism for which we can easily measure 0.001 eV.
All these energies are easily measurable, and they do not include some kind of “information” or “algorithm” pushing things around.
Suppose that, instead of scaling ourselves down to mingle among atoms and molecules, we scale up the forces and energies down there to what they would have to be at our macroscopic level. Here is an easy way to do it.
The ratio of the electrical force to the gravitational force on two protons is 1.2 x 10^36.
Suppose the electrical forces among kilogram-sized masses in our macroscopic world were 1.2 x 10^36 times greater than the gravitational forces acting among them. What would that mean for the energies of interaction? Potential energies and kinetic energies would scale up by the same factor.
So a binding energy (potential well depth) of 1 eV scales up to 1.2 x 10^36 eV which is 1.9 x 10^17 joules, which is equivalent to 47 megatons of TNT. In other words, to break the bonds of interaction would require 47 megatons of TNT.
The typical bonds in solids scale up to 4.7 megatons of TNT. The typical bonds in liquid water and in living organisms scale up to 0.47 megatons of TNT.
Down at the atomic and molecular level, electromagnetic interactions are not trivial. Things like “information” and “algorithms” would have to push atoms and molecules around at levels greater that the electromagnetic interactions.