Can Evolution Learn?

Thanks for posting this Astyanax. I read the full article and found it fascinating.

Other than that, I don't have much to add at the moment.

a reply to: Astyanax

I guess I get antsy when people anthropomorphize things. Evolution is not capable of learning, as it is a process, not a thinking entity. That paper seems to be almost trying to say "I am not saying God did it but .... God"

a reply to: Noinden

The title `How Can Evolution Learn?' is that of the original paper, as published in Trends in Ecology & Evolution.

Some mental flexibility is called for here. What is the verb for what a feedback-based self-adjusting and adapting process is doing? I think 'learning from experience' captures it nicely.

The grammatical category `abstract noun' exists to permit exactly this type of conceptualization. If you wish to abolish 'evolution' as a noun, you must also abolish truth, justice and happiness.

a reply to: Astyanax

Seems like a no brainer that this would be the way evolution works. Evolution (and for that matter, the genome) is plastic, responsive, "trainable", and yes, intelligent.

originally posted by: Astyanax
This hypothesis introduces a teleological factor into the model of evolution by natural selection.

Doubtful this is wishful thinking on your part, although I could be wrong. So I'd be interested to understand why this was your takeaway from the study/article.

As I see it - learning implies intelligence.

Also, as the evolvability of a genome increases, doesn't the dependence on natural selection to produce a beneficial trait decrease?

originally posted by: Astyanax
Of course, the authors of the paper are not suggesting that the hypothetical bias towards producing useful rather than deleterious mutations was God-given, but that it is itself a product of evolution...

which from your takeaway is now potentially a step closer to being considered a teleological process...

originally posted by: Astyanax
However, Creationists may be tempted to embrace the idea that life has a ‘designed’ preference towards evolving beneficial adaptations.

If Watson's hypothesis gains more traction then yes, life would seemingly have a designed preference towards beneficial adaptations - thanks to an intelligent process that learns and becomes better as it goes.

Thanks for the post. s&f

originally posted by: Noinden
a reply to: Astyanax

Evolution is not capable of learning, as it is a process, not a thinking entity. That paper seems to be almost trying to say "I am not saying God did it but .... God

No, that's not what the paper is almost trying to say, but you are entitled to your opinion of course. Perhaps that's just your paranoia poking at you.

Also, why does learning have to be completed by a "thinking entity"? You may have to define what you mean by that because AI, computers and heck, even plants can learn. But are they necessarily thinking?

a reply to: PhotonEffect

Seems like a no brainer that this would be the way evolution works.

From the abstract:

A simple analogy between learning and evolution is common and intuitive. But recently, work demonstrating a deeper unification has been expanding rapidly. Formal equivalences have been shown between learning and evolution in several different scenarios...

I won’t pretend to understand the specific examples of the equivalences that are then listed, but there is clearly more to the point they’re making than ‘every successful adaptation increases the likelihood of further successful adaptations.’

I'd be interested to understand why this was your takeaway from the study/article.

George Bernard Shaw was among those who, in his time, championed a concept known as ‘intelligent evolution’. According to this, Life (another abstract noun) was possessed of an intrinsic urge to attain ‘higher’ and ‘higher’ forms through evolution. Although no actual deities were involved, nothing could be more teleological. But evolution doesn’t actually work like that, and claims for intelligent evolution invariably drifted into the realm of metaphysics anyway, so the whole thing ended up in the Bin of Discarded Ideas.

While the paper we’re discussing is no exercise in metaphysics, it does posit a discernible end that evolutionary processes have in view: adapting successfully to ambient conditions. That’s where the teleology comes in. Evolution may be showing signs of having an object. This is no part of the Modern Synthesis. This is from out of left field.

as the evolvability of a genome increases, doesn't the dependence on natural selection to produce a beneficial trait decrease?

Wut pliss is meaning ‘evolvability’?

a reply to: Noinden

It expresses both awareness (byway of the use of knowledge) and will (being goal orientated and use of awareness).

There's no way around it, so you're right in your assessment there.

But the part that I think you do have wrong is in you thinking they're wrong. (It's not "anthropomorphism": actual will and awareness are somehow involved.)

a reply to: Astyanax

originally posted by: Astyanax

A simple analogy between learning and evolution is common and intuitive. But recently, work demonstrating a deeper unification has been expanding rapidly. Formal equivalences have been shown between learning and evolution in several different scenarios...

I won’t pretend to understand the specific examples of the equivalences that are then listed, but there is clearly more to the point they’re making than ‘every successful adaptation increases the likelihood of further successful adaptations.’

I won't either, but it does seem apparent that a connection is trying to be made between evolution and it's ability, as a process, to learn. Or the ability of a genome to learn. Either way, it all sounds very provocative. Were you able to gain access to the full article by chance?

originally posted by: Astyanax
While the paper we’re discussing is no exercise in metaphysics, it does posit a discernible end that evolutionary processes have in view: adapting successfully to ambient conditions. That’s where the teleology comes in. Evolution may be showing signs of having an object. This is no part of the Modern Synthesis. This is from out of left field.

Evolution with a purpose... hmm

originally posted by: Astyanax
Wut pliss is meaning ‘evolvability’?

I gather, from what I've read and the manner in which evolvability is defined in the article, that it refers to how well a genome/organism/species evolves. Watson seems to be saying that genomes in general are getting better at evolving, or in other words producing phenotypes that are better at adapting to a changing environment.

Adapting is defined within the context of natural selection, but then the article says this:

Like a child, a neural network cannot make the connection instantly, but rather must be trained over time. That training is complicated, but in essence it involves changing the strengths of the connections between the virtual "neurons". Each time, this improves the result, until the whole network can reliably output the desired answer: in our example, that the funny symbols on the page ("hello") equals the word "hello". Now the computer "knows" what you have written.
A similar thing happens in nature, Watson believes. An evolvable species would "output" a trait just right for a given environment.

Perhaps this is the teleological aspect (if not somewhat Lamarkian).

But if a genome is able to 'output' a trait that is just right for the organism within its current environment, then where does NS come into play? If this ability to output beneficial traits is getting better, as Watson is suggesting, then it sounds like the phenotype is being given exactly what it needs without the need of a trial and error process. Selection does not play a role, or does it?

He then goes on to mention atavisms, which are:

..., traits that have disappeared phenotypically but do not necessarily disappear from an organism's DNA. The gene sequence often remains, but is inactive. Such an unused gene may remain in the genome for many generations.[5]

But this isn't what we're told about how natural selection works - is it? Isn't natural selection supposed to weed out useless traits/alleles for good?

Everything is in a state of decay. Second law of thermodynamics. Evolution contradicts this.

originally posted by: jjkenobi
Everything is in a state of decay. Second law of thermodynamics. Evolution contradicts this.

Not even remotely true. The 2nd law of Thermodynamics applies only to a closed system. Neither the earth nor its inhabitants are part of a closed system.

a reply to: Astyanax

I am aware, I am just pointing out my issues with the title. This is not about "metnal flexibility". So you can drop that argument. Its about a poor choice of words, given the people who will latch onto the title, and not read the content.

a reply to: Bleeeeep

As we've not interacted before, be aware I'm a scientist, one who has a background which includes Bioinformatics, which is one of the sciences you can study evolution with. I'm also a very spiritual individual. Just not Abrahamic. SO the answer "god" is always "which one" with me

I've seen zero evidence of scientific principals having awareness. Otherwise (say) the SN1 and Sn2 mechanisms would have a sense of humor....

a reply to: Noinden

"As we've not interacted before, be aware?" Your personage means nothing to me. Saying you're qualified means nothing if you aren't. And similarly, saying you understand means nothing if you don't. You think you're trained in the proper school of thought, like you have the formula, but you don't. You don't have the answers - you have delusions of grandeur and you need to check that if you want to think clearly enough to understand.

You good?

Now follow along...

Does being aware mean being aware of everything? No.

Does being aware mean being aware of all of your reactions? No.

Does being aware mean being self-aware (consciously aware)? No.

Is there a difference between being aware and being self-aware? Yes!

So, what does it take to classify something as being aware, not self-aware, but aware? Seriously think about that. Wouldn't said classification look like reactions without self-expression (without free will) and be, for the most part, without deviation (like chemical reactions)? I think so. So why do we think otherwise? Why do we even think that there can be a reaction to anything at all without an awareness to react to it? It's because of materialism, right? People have told us that reality is a dead mechanical thing, but today, we're moving away from that - today, Newtonian physics is dead, and reality is a thing of living will. And that's what you're seeing in this study: people are coming to terms with the fact that awareness doesn't begin and end with the human brain. (Remember when we used to question whether or not animals were conscious? That's how ridiculous your school of thought is today.)


You don't have to be aware of everything to show awareness - you don't even need to be consciously aware (self-aware) to show awareness - the only thing you really need is a force of orientation or self-organization. Where there is information being processed, there's awareness.

Oh almost forgot:

a reply to: Astyanax

I was surprised that the authors of the article in Trends in Ecology and Evolution didn't reference this paper in their literature search. The theme is essentially the same, however, the approach is entirely different. I prefer their approach because they were able to simulate the interaction between learning and evolution. Below are a few excerpts from the 1987 paper by Hinton and Nowlan, Computer Science Department, Carnegie-Mellon University.

If anyone wants a copy of both papers, I can make the PDFs available - just send me a PM.

a reply to: LSU0408

sorry

a reply to: Bleeeeep

ou don't have the answers - you have delusions of grandeur and you need to check that if you want to think clearly enough to understand.

I have interacted quite sufficiently with Noinden to know that his or her claims are valid. Your characterization of him or her is utterly false.

I have also interacted quite sufficiently with you to form a pretty good opinion of the value of your contributions. You’ve had your say. Nobody’s biting. If you have nothing more to add that is on topic, please go away.

a reply to: Astyanax

You trying to vouch for his ego with your ego, to counter my pointing out that his ego can not rationally be used as proof of his claim against intelligent evolution, is like wanting to go double or nothing on your horse because you're too stupid to know when your horse is dead. Do you want me to beat a dead horse?

Okay then: Your claim is that your ego (your opinion) is evidence for his claim (that his ego is proof of his claim), but like I told him: personage means absolutely nothing - it plays no part in the truth of the matter and so your argument is not based in valid reasoning or objective truth - it's just based in stupidity, which actually detracts from your credibility.

And with that said, I'd like to echoe your request: if you have nothing more to add that is on topic, please go away. And yes, I mean it. It was nice of you to bring the topic up but clearly it is beyond you - so please, don't get in the way of people who can think and discuss the topic rationally.

Not sure if this is considered evolution, or simply adaptation.

An increasing number of African elephants are now born tuskless because poachers have consistently targeted animals with the best ivory over decades, fundamentally altering the gene pool.

In some areas 98 per cent of female elephants now have no tusks, researchers have said, compared to between two and six per cent born tuskless on average in the past.

Almost a third of Africa’s elephants have been illegally slaughtered by poachers in the past ten years to meet demand for ivory in Asia, where there is still a booming trade in the material, particularly in China.

www.independent.co.uk...

a reply to: angeldoll

That might be natural selection, in that, the only thing left to breed were short tusked/no tusked elephants. But then again, the trait's (short tusked/no tusked trait) (re)emergence might be a sign of intelligence (think: order must preexist causality so where did the order come from? was it random order or was it conceived, as by an awareness, order).

The ordering, and its method, is what's in question. Of the seemingly infinite number of random mutations, why short tusks and not orange ears and a pink trunk? You can say, "well they would die off." But really, they are never born. So, where's all the elephants with octopus suckers and butterfly wings? They don't exist and "random" seems a lot more like what you would expect if you were expecting goal orientation.

a reply to: peter vlar

This has to be one of the most willfully ignorant and uneducated bits of poo I've seen regarding the modern evolution synthesis in quite awhile. That's impressive considering some of the garbage posted on ATS daily.

You are just encouraging him/her/it, Peter.

Try not to feed the trolls.