Weasel and other genetic algorithms have been falsely advertised as proof of concept of Darwinian evolution in nature. Use of the term “natural selection” is false advertising because what Darwin and evolutionary biologists claim is natural is actually NOT natural.
I had an offline exchange with EricMH as I’m thinking of contributing to the Blyth Communication journal regarding computational models to highlight Darwin’s false advertising of what natural selection is really capable of doing. Whether my ideas end up in the Blyth journal is a separate issue from the ideas themselves. I’m hoping EricMH will weigh in with some ideas too.
EricMH suggested some other ideas offline, and I asked if it would be ok with him to discuss this at TSZ to get editorial feedback. I suggested we avoid use of CSI but instead using things like common sense to dismantle Avida, Weasel, Tierra, and other GA’s used as “evidence” that natural selection actually provided the complexity we see in biology today
Bill Dembski attempted to do the what I’m trying to do with pure math, but I instead decided to use common sense.
I hope to post some ideas here in this OP, but also in the comment section. If people object to the simplicity of the models, they’ll have to justify why my models are less valid than their GA’s or computer models like Avida, Tierra, Weasel or whatever. The point is to highlight the absurdity of assuming that because something is a called a “GA” that it actually models the physical genetic “algorithms” in nature or that there is no need of miracles to make evolution happen.
So for starters, here are some ideas:
Frozen in Space Weasel — there is no requirement in nature that an organism be provided temperatures and atmospheric pressures appropriate for life and genetic inheritance to take place. In fact, MOST of the universe is inhospitable, and for that matter, it takes fine tuning for life to emerge. It certainly is NOT natural (in the sense of expectation from first principles) that life should be sustainable. It is an exceptional event. Why should we assume a GA will run naturally?
DOA Weasel — well, without the origin of life, there is no Genetic Algorithm in real life. Evolutionary theory, as always, needs a miracle to make it work, and in this case a miracle to even get it started.
Global Warming Weasel — because of the Faint Young Sun Paradox, the Earth should have been a frozen ice ball till now, and there shouldn’t have even been a Cambrian explosion. Supposedly green house gases were miraculously fine tuned to allow just the right amount of global warming to counter balance the Faint Young Sun. The global warming gases miraculously diminished just at the right time and in the right amounts over billions of years to enable habitable temperatures to enable life to evolve. As usual, evolution needs a miracle to make it work.
See: The Young Faint Sun Paradox
Devolving Weasel — this weasel gets simpler and simpler as natural selection destroys its complexity. It is even LESS likely than random chance to arrive at the magic phrase, “methinks it is like a weasel.” This seems the natural direction of evolution, not toward complexity. Thus again, evolution needs a miracle to make it work, because selection toward extravagance and complexity from a starting point of simplicity is not indicated by theory nor experiment. Thus evolution again needs miracles to make it work.
Thwarted by IC Weasel — This is a weasel that can’t evolve because irreducibly complex systems don’t provide a feedback path, or worse, a NEGATIVE feedback path toward evolving complexity. To quote Michael Lynch, ” many genomic features could not have emerged without a near-complete disengagement of the power of natural selection.” I had a conversation with Robert Marks about this, but then we ran out of time to finish out the idea…
Endagnered species Weasel — this weasel dies out BECAUSE natural selection as its habitat is taken away by other creatures. So, for natural selection to enable diversity, natural selection has to be absent!
So we may add a few more ideas, and then Python code, or whatever code to illustrate these ideas.
It would probably help, imho, if some humorous sounds and vizualizations were added to the python code to drive the point home with the above algorithms. The point is to mock the absurdity of using GAs as some sort of evidence in favor of Darwinian evolution in nature. Bill Dembski attempted this in several publications including No Free Lunch using piles math. But I feel a more common sense approach will resonate with a general audience as well as a scientific audience.
Because I feel these ideas more appropriate reflect the reality of how miraculously natural selection must work in order to evolve complexity, and to mock programs like Avida, Tierra, Weasel, I call these the REAL WEASEL family of Genetic Algorithms.
Abiogenesis Evolution.
Also, https://www.researchgate.net/publication/224301917_Irreducible_complexity_in_a_genetic_algorithm
I propose “Salvolution”, which doesn’t change even though the information environment moves forward at an exponential rate.
Bonus: “Endagnered species Weasel — this weasel dies out BECAUSE natural selection as its habitat is taken away by other creatures. So, for natural selection to enable diversity, natural selection has to be absent!”
If only people had researched extinction rates! Dagner Will Robinson!
Sal,
Everyone who knows anything about genetic algorithms understands that their success depends on the shape of the fitness landscapes they navigate.
If you want to show that evolution is impossible, you’ll need to demonstrate that the fitness landscapes have the wrong shapes. Good luck with that.
Richard and Keiths, thanks for the comments. Let me think on them.
And thanks in advance to all who participate in this discussion.
It’s not a need for discussion that promoted this. It’s your cognitive dissonance manifesting itself.
https://en.wikipedia.org/wiki/Consilience
They are less valid because they don’t exist.
Avidia, Tierra I can go and download. Weasel I can write. Can I say the same for the pieces of shit you sprayed into the OP? No I can’t. That’s why they are less valid. Got it?
And I thought J-Mac had found it but no, it seems that under the bottom of the barrel there’s a sub-basement and in there we find this turgid mess.
No matter what, it’s evidence for you.
If the universe was 100% hospitable that would be evidence for design. That it’s almost the inverse, also evidence for design.
cowed said that when he was investigating biology and found that some bit of some system was controlled by another system, that was what convinced him that ID was real. I asked him was there anything else that he could have discovered at that point that would have not led him down that path. He ignores the question to this day.
So, a universe where we can see back to the beginning of time and still see no evidence of any other life is evidence for the design of our life. The hubris.
So you’re going to build a few models that fail, and then blame Darwin for your own failure. Okay then. Have fun.
Blasted REAL Natural selection selecting for pot smokers over weasels!
https://www.latimes.com/local/lanow/la-me-marijuana-threatened-species-20180622-story.html
Where do GA’s model this scenario where one species causes the extinction of another, and hence there is no further evolution.
How about a cataclysm like a meteor strike or other NATURAL disasters!
One could just as well say the GAs like Avida are the EXCEPTION rather than the rule for REAL evolution.
The same mistake as always.
You are trying to prove that GAs don’t create complexity. Well, of course they don’t. The complexity is already there in the environment. The GAs just adapt to the already existing complexity.
When can we expect that common sense to show up? Right now, it seems to be missing.
dazz, to Sal:
That pretty much sums it up.
I’ve never seen such kind of advertisement. Genetic algorithms show that random variation and selection lead to interesting things. To solutions where engineers don’t know how to approach some problem, to semi-optimizations, etc. But “proof of concept of Darwinian evolution in nature”? That sounds like the person who wrote the OP mistakes concepts for referents.
Which is why I’ve never seen such kind of “advertisement.” Selection in genetic algorithms is not “natural.” What imbeciles like Salvador miss, constantly, is that the combination of a blind process for producing variation, with selection for features, can build complexity. It does, and there’s no way around it.
And here is where I stop. Eric contradicts his own statements and claims within single sentences. Then he contradicts another thing he said in the next comment. You try and make him aware of his deep philosophical and scientific messes, and he never attempts to give an answer, let alone to try and correct his misconceptions.
An illiterate buffoon, like Salvador, combining “forces” with a walking contradiction. The magical being in the sky surely knows how to choose “his” defenders. It’s so ridiculous that it makes me suspect that “he” doesn’t exist.
It is more accurate, unlike Dembski and others, to say I’m trying to show
1. GA’s like Avida, Weasel, etc. don’t model what actually happens naturally
2. the environment doesn’t not naturally create improbable structures, as in structures far from random expectation, and that selection isn’t naturally expected to do better than random expectation either. This is what Dembski attempted to show mathematically with No Free Lunch, but which is actually easier to describe with actual known issues in biology.
What actually happens naturally, according to you, are miracles. Model them?
Would you like some dressing with your word salad?
You have a double negative, so it isn’t clear what you are saying.
Yes, the environment does create highly improbable structures. You just happen to not notice that.
Ah, but you could argue that all structure so created is highly improbable. I.E. solar system setup goldilocks ratio etc.
If only there was some way to distinguish intentional design from potentially change happen stance. A reliable way to achieve that would indeed be a boon!
Is anyone aware of such a thing? It’d be like some sort of sieve, where we trawl facts and physics for possible pathways for the way things could have happened ruling out what could be explained by mere chance alone perhaps? Something like that? Where we’d filter our explanations until we were left with the conclusion – design!
I bet some of our ID friends know if such a thing exists? Such an filter of explanations?
First, you model how the first functioning cell aroused:
First, you model how enzymes aroused but to accomplish this you need energy from ATP. So, you model how ATP aroused but to accomplish this you need enzymes.
So, you need a miracle to model this fundamental step in the origins of the first cell.
Now, you model how DNA aroused but to accomplish this you need enzymes. So, you need another miracle to model the next fundamental step in the origins of the cell.
But for those molecules to even function they need to be protected by a cell membrane but to accomplish this fundamental step you need proteins to make the cell membrane. But proteins can only be made by a cell. So, you need another miracle to model this fundamental step in the origins of the first cell.
But this is not the end of it, a functioning cell needs thousands of proteins, so to model how the first cell arose, you need thousands of miraculous…
But even if you get all the necessary components for the first cell and put them together somehow by a miracles you need another great miracle: you need to make an inanimate matter animate…
So, you can try to model that…When you have done it all, please report back your ignorance…
Dawkins wrote a computer program to simulate a monkey at a typewriter. He referred to the program as the monkey/Shakespeare model of cumulative selection, and explained that it was not a model of cumulative natural selection. One of his main points, illustrated by the model, was that cumulative selection is not necessarily a biological process.
Thanks Tom.
Typo alert aroused = arose
Sorry for my bad grammar. Thank you for pointing out my error.
The environment does not create highly improbable structures (improbable as in far from physical expectation).
Any arbitrary arrangement of dominos might be highly improbable, however, the dominos below are improbable as in “far from physical expectation” from arbitrary conditions and random processes.
By way of extension, there are arrangements of molecules that are in a configuration far from random expectation. And further, there are configurations far from expectation of random processes and/or not achievable by mutation and selection.
Unless most mutations are not truly random…
“Weasel and other genetic algorithms have been falsely claimed by creationists as intended to prove the concept of Darwinian evolution in nature.”
Fixed that for you.
The Weasel simulation by Dawkins was intended to do one thing and one thing only. Creationist debaters had frequently claimed that evolutionary biologists put forward a theory of change that is purely at random. The implication they hoped that their audience would take from this is that evolutionary biologists explained adaptations by random changes that are not biased in the direction of adaptation. The audience would then, of course, reject this as patently absurd.
That is of course a deeply dishonest argument. So Richard Dawkins made up a teaching example, where selection biases the changes so that they tend to be toward adaptation, and does so by selecting the offspring closest to being well-adapted. And his example shows cumulative selection, with each generation starting from the results of the previous one. Real organisms of course must do that, as they are the offspring of the previous generation.
Dawkins was then easily able to show that the Weasel became well-adapted vastly faster than pure random wandering.
The Weasel has been so effective a teaching example that creationists are generally furiously upset with it. They have then invented all sorts of mischaracterizations of what Dawkins was up to, and each time triumphantly declared that his Weasel failed to show that or to do that.
Here’s to the Weasel, still infuriating creationists after 30 years!
Sure, it does. Every snowflake is different. So every snowflake is highly improbable.
Well, there’s your problem. You expect every snowflake to be different, and therefore this does not violate your expectation. So you disregard it.
Joe:
And to all the Elmer Fudds, including Sal (and now evidently Eric), who have twied and faiwed to catch that wascally weasel over the years.
And that’s how the concept of non-random natural selection was born…
From this moment on, Darwinists could make any claim they wanted because if natural selection is non-random it may as well be omnipotent…
Praise the Lord!
??? I can make non-random selections much more consistently and accurately than natural selection. Does this make me super-omnipotent? News to me!
J-Mac,
Keep shooting yourself in the foot J-Mac. It’s amusing.
Entropy:
But it gets blood all over the carpet.
Ummmm, no. The idea of non-random natural selection was invented by Charles Darwin and Alfred Wallace and published in 1859.
Why do you continually say things that are so easily shown to be nonsensical?
Tom English,
Joe Felsenstein,
Sal knows this full well, but remember his goal is to make a quick buck off of some uneducated creationists. Therefore, not lying about Dawkins is simply not an option.
Hear hear!
I was “Blue Lotus” at the time on UD, and had a lot to do with helping getting “Latchgate” started. Fantastic times.
And then we found the video and KF still doubled down anyway.
Most selection is NOT cumulative, that’s Dawkins first mistake and misrepresentation, and the error of most evolutionary biologists. The only place it is NET cumulative is in the imaginations of evolutionary biologists, not actual experiments and observations.
Improbable as in far from expectation! If every snowflake is different, it would be violation of expectation if 100 snowflakes in close proximity in time and space were identical. That would be improbable in terms of violation of expectation, correct?
How does this relate to REAL GENETIC algorithms? For genetic algorithms in nature to emerge, there must be an origin of life event to get the whole thing going. Well what is improbable and far from expectation is:
Homochiral amino acids for long stretches in initial proteins which enable the existence of alpha helices that make functional proteins possible.
Alpha peptide bonds connecting all the amino acids in linear fashion in the initial proteins when there are a variety of other bonds.
There are more sophisticated examples of violation of expectation such as evolution of ATP Synthase and Helicase, but those listed above are some examples.
James tour listed other problems that involve violation of expectation. This violation of statistical expectation can also be expressed in terms of violation of the Gibbs free energy.
I have NOT failed. Here is my proof:
Bwahaha!
Joe Felsenstein,
It showed a mind was used to create the algorithm that could find a small sequence from a random search with selection from that search.
It may have upset creationists but in the long term since no one has navigated a sequence of a Weasel length without a human devised target the exercise turns out to support the intelligent design claims.
keiths:
Elmer/Sal:
Evidently, we need to teach Elmer/Sal the difference between a wabbit and a weasel.
You mean most individuals do not start out with genotypes that come from genes transmitted by their parents?
That’s what makes for the “cumulative”. It’s not some mysterious property involving long-term cosmic resonance. It’s that, before that generation’s selection, each generation starts from where its parents were.
colewd,
That’s silly. The target doesn’t need to be “human devised”.
Thank you for reading and responding, but that is NOT what I meant.
Creatures tend to lose integrated function over time, not accumulate it.
The first example of this loss of function is outright extinction.
The next example of this is REDUCTIVE evolution. This is well observed and documented.
The frequency of reductive evolution and extinction of integrated function is seen more frequently than construction of new integrated function. For example, there is gene loss in bacteria and other parasites. It’s easier to lose the ability to fly than it is to acquire it from scratch. Same for vision, etc.
Hence the notion of cumulative (as in accumulating more and more integrated systems), is not in evidence by direct observation nor experiment. If one wants to argue junk is accumulated, then that’s fine, but that is not the sense of what I mean by cumulative.
The only place there is NET accumulation of complex integrated function is in the imagination of evolutionary biologists, but not in direct observation and experiment.
Were it not for horizontal gene transfer, selection would cause widespread loss of genes in bacteria, for example:
https://academic.oup.com/gbe/article/7/8/2173/557455
So if we were to model this in the REAL WEASEL family of Genetic Algorithms, we might start out with: “METHINKS IT IS LIKE A WEASEL” and then it evolves to something like: “ME IS A WEASEL”. Wouldn’t that at least model an experimentally an observationaly confirmed example of selection at work versus one that was just rooted in Dawkins and Darwin’s fantasized notions of how natural selection actually behaves.
See, we don’t need CSI and mathiness to show this stuff…it’s right there in the data.
My weasel is better than Dawkins’. 🙂
Gene gain… gene gain 😜
Thanks for responding.
“ME IS A WEASEL”
keiths,
Maybe depending how you define “human devised” but me thinks a human is somewhere in the flow chart of random letters and “Me thinks it is like a Weasel”.
I’m not sure I follow this argument. Are you saying that the very first bird didn’t evolve flying ability by taking advantage of other inherited features (perhaps used for other things like body temperature regulation), buy rather was poofed whole cloth into existence by your creator.
And are you saying that all of the hundreds of species of birds today, from hummingbirds to condors, are cumulatively degenerate forms of that first poofed bird, losing their ability to fly (even though some species rarely even land). Interestingly, the visual abilities of many of these birds has degenerated into something far superior to our own ability.
I think your claims would beguile more readers if nobody ever bothered to look around at the biosphere and notice that the exact opposite of your claim, has actually been happening. Bummer when observation refutes preference, even when that preference is buttressed by utterly ineducable blind faith.
“Out of 120,000 fertilized eggs of the green frog only two individuals survive. Are we to conclude that these two frogs out of 120,000 were selected by nature because they were the fittest ones; or rather – as Cuenot said – that natural selection is nothing but blind mortality which selects nothing at all?”
I think you got me confused with the Darwin boys…😉
This is a conceptual mess Bill. The search is not random, the search includes all of it. The random “mutations”, the selection, and the reproduction.
Then I don’t understand why you insist on talking about the “mind” that wrote the program, rather than about what the algorithm does. Why the change of focus? Did you notice that “mutations” were random? Did you notice that the elements that make it work, “despite” the random nature of the “mutations,” are selection and reproduction?
Are you insinuating that you accept evolution, but that it is some “algorithm” written by a magical being in the sky?
If you accept evolution as an algorithm written by a magical being in the sky, because Dawkins wrote a program that shows that random variation with selection and reproduction can produce meaningful text, then you’ve been contradicting yourself for quite a while arguing against the possibility that evolutionary phenomena have been producing functional molecules.
You cannot have it both ways.