Intelligent Evolution?

[quixotecoyote]

Oh, c'mon guys, give me a break. You're not trying to have us all believe that the hypothetical extensions of the OP analogy as valid similies for biological evolution are too far removed from the OP analogy to invalidate it, are you? Hell, we can run the extended analogy and examples by the IDiots on the basis that we can actually and physically emulate them in practice to prove their validity, and have the IDiots on the first train home tomorrow. Just go find some disgruntled fruit pickers and start making ladders for them without consulting!

No you can't. You cannot show that the majority of technological development is like your extended analogy because of the simple fact that it isn't.


Some statues of people are made of marble.
That doesn't means that describing marble as looking like people is at all helpful.

You take a category, define an extremely rare hypothetical subset of that category, then try to pass that off as representative of the category. It simply does not work as an analogy to technological development. If you actually created your automaton scenario it would go beyond an excellent analogy and into the category of an excellent model. Unfortunately none of that helps the OP as that model would still not be representative of technological development.

I'll tell you what I think. I don't think you two and others like you really know what you believe! You simply twist and contort your watery argument away from any insinuation that you mistakenly think might curry favour from IDiots; insinuations which, if properly understood, actually counter the ID argument.

Again you stoop to creationist smears.

You are a liar.

I had some respect for you in our earlier conversation, but now you're back to straight out lying.

I am not a creationist.

I am not an ID'er

You are a liar for saying I am.

Liar.

 

[jimbob]

Quote:
I'll tell you what I think. I don't think you two and others like you really know what you believe! You simply twist and contort your watery argument away from any insinuation that you mistakenly think might curry favour from IDiots; insinuations which, if properly understood, actually counter the ID argument.
Again you stoop to creationist smears.

You are a liar.

I had some respect for you in our earlier conversation, but now you're back to straight out lying.

I am not a creationist.

I am not an ID'er

You are a liar for saying I am.

quixotecoyote, I may be wrong in reading Southwind's posts but I think southwind is saying that we are mistaken in asserting that IDers like the analogy. It is hard to understand, but it is hard, as IDers use this very analogy themselves.

Nowhere do I see IDers say "Intelligent design works because it is unlike evolution"

They say "evolution is just like technical evcolution, which is actually intelligent design"

ETA:

Southwind, what is "watery and unllclear anbout our posts"?

The processes are fundamentally different, yet the analogy is attractive for the wrong reason, and doesn't advance understanding, but could hinder it.

that is why I am arguing againse the analogy, especially as posted in the OP.

 

[jimbob]

Southwind, do you accept that even wioth your "let the maket decide" evolutionary algorithm story, each Successfull variant is chosen by an individual customer, who is an intelligent agent.

The selection is thus intwo parts,: Firstly by many different intelligent agents, and then according to other selection criteria to cull out the failures. In any system with finite resources the second is also vital. Every realistic system as finite resources.

 

[quixotecoyote]

quixotecoyote, I may be wrong in reading Southwind's posts but I think southwind is saying that we are mistaken in asserting that IDers like the analogy. It is hard to understand, but it is hard, as IDers use this very analogy themselves.

You're right.


Southwind, I apologize. When you said.

You simply twist and contort your watery argument away from any insinuation that you mistakenly think might curry favour from IDiots

I read it as if you were saying we were currying favor from ID'ers. It appears on second reading that you're saying we are doing so, but not intentionally.

I still think you're mistaken, but I shouldn't have accused you of parroting articulett's line so quickly.

 

[fishkr]

Random variation is irrelevant to Darwinian Evolution.

I'm guessing he meant "random mutation".

 

[fishkr]

I am not arguing that it is couldn't be described as bottom-up.

QUOTE]

Yes. But isn't this the crux of the matter at hand? It is clear to us that ET is mechanistic and bottom-up, but in the course of the present discussion the bottom-up aspect of TE is paramount to the question of how the analogy may be used by IDiots.


Interesting that our fields overlap. Six7's too, but he's being quiet about it.

Now I'm going to take a stab at what we disaggree on:

I think there is no "solution space" at all, at least not one that is discrete from any other space. I realize that this is just a figue of speech, and forgive me for pirating your term, which I take to mean the context within which thought and problem-solving stimuli and responses occur. I hope I'm maintaining your meaning, but in my mind a "solution space" is an intellectual contrivance that we all participate in because it's usefull and makes sense to us as individual organisms, but with a bow to Apathia and A.N. Whitehead, I think this is a process/object problem, and illusiory, and key to the disagreement.

The tech environment is a seamless continuum, past and present, which is part of a larger seamless physical continuum which includes the natural environment, past and present. You can chose to view the TE process in terms of egos laboring over solutions to engineering problems, or you can focus on the engineering "problem" itself as part of a larger continuum. In this way the continuous "problem set" of life, and the "solution space" you speak of can be thought of as one process moving through time, just as we can look back into the fossil record and see snapshots of biological "solutions" reduced to stone, or through the Patent Office as snapshots of TE.

IMHO the key difference between ET and TE is the inclusion of human consciousness. Being self-aware, we are different from giraffes reaching for apples . We can reflect, analyse, invent. Our inventions are constrained by context, always, and our impulse to invent is as natural as reaching for an apple.

Why should we think of our consciousness as being apart from the natural environment? Isn't that where it came from? Isn't that where it lives?

The tensile strength of a steel cable is a condition of our environment that a lift designer would respond to if he wanted to make a functional lift that didn't kill people. If he didn't do that he would be selected out of the meme pool in favor of someone else. If he specified too large a cable, he would be selected out by the owners of the lift company because it would be heavier and more costly than needed. I see nothing in this process other than survival of the fittest meme, rather than gene.

If TE turns out to be a bottom-up process that can be shown to shadow or mirror ET, isn't it more important to represent our best understanding of reality than to worry about politics?

 

[fishkr]

Southwind, do you accept that even wioth your "let the maket decide" evolutionary algorithm story, each Successfull variant is chosen by an individual customer, who is an intelligent agent.

.

Are viruses "intelligent agents" because they respond to a protein key?

 

[six7s]

I am not arguing that it is couldn't be described as bottom-up.

Yes. But isn't this the crux of the matter at hand? It is clear to us that ET is mechanistic and bottom-up, but in the course of the present discussion the bottom-up aspect of TE is paramount to the question of how the analogy may be used by IDiots.

Interesting that our fields overlap. Six7's too, but he's being quiet about it.

Indeed

Are viruses "intelligent agents" because they respond to a protein key?

And I am likely to remain (relatively) quiet when the discussion is considering the concept of 'intelligence' - a semantic can-of-worms if ever there was one

"Alright," said Deep Thought. "The Answer to the Great Question …"

"Yes …!"

"Of Life, the Universe and Everything …" said Deep Thought.

"Yes …!"

"Is …" said Deep Thought, and paused.

"Yes …!"

"Is …"

"Yes …!!!…?"

"Forty-two," said Deep Thought, with infinite majesty and calm.

"Forty-two!" yelled Loonquawl. "Is that all you've got to show for seven and a half million years' work?"

"I checked it very thoroughly," said the computer, "and that quite definitely is the answer. I think the problem, to be quite honest with you, is that you've never actually known what the question is."

Ring any bells?

 

[Southwind17]

The problem is that if one's business is making ladders for people, one would not make ladders without first consulting the people on what they needed from a ladder. Similarly, while real engineers may not start with all the answers to the questions the want to address, they are not completely ignorant of how to make things work as Ollie was with his little electronics kit. As I have said before, all these examples prove is that by removing the essential elements of development and replacing them with the essential elements of evolution you can transform development into evolution, a trivial result in itself.

Mijo, you really do need reminding of what the OP analogy set out to do. Allow me. It attempts to show that seeming irreducible complexity can and does arise from certain technological developments following an iterative process which, at its basic level, dilutes the 'intelligence' factor so much so that it can essentially be discounted. At worst, it serves to contextualize, and possibly quantify, the degree of 'intelligence' borne by the IDiots' make-believe creator to that of mere mortals. If the OP analogy doesn't do it for you, then what's wrong with adopting the hypothetical extensions of the OP analogy instead? Indeed, what the extensions do is remove intelligence completely and show that seeming irreducible complexity will still arise. Not only is it easy to envisage the analogy extensions but they can easily be modelled, to prove their validity. So, mijo, let's drop the OP analogy and go with the extensions where 'intelligence' is replaced with randomness. We've then set up the perfect model for proving that 'intelligence' is unnecessary, which was the whole purpose of the OP.

But you'll never be successful in the real estate business if you refuse to acknowledge that a Tudor-style house is different from a Victorian-style house because they are both houses. People will also not take you seriously if you insist that they are the same because you can transfer the essential design elements from one to the other.

Oh, so the blue house has now become a Tudor-style house, and the red house a Victorian-style. You see mijo, this is a perfect example of this:

I'll tell you what I think. I don't think you two and others like you really know what you believe! You simply twist and contort your watery argument ...

 

[Southwind17]

Most analogies, though, have more that one point of correspondence, because almost every pair of objects has one point of correspondence (e.g., raven and writing desks both have quills); therefore having one point of correspondence doesn't necessarily make a good analogy.

So what is the only single point of correspondence that you believe technological development displays (incremental change over time, presumably), and how many other potential points of reference do you consider there are (natural selection, i.e. survival of the fittest, presumably)? That makes two.

So why do you consider the retail environment as NOT analogous to the 'biological environment' where natural selection occurs? Consumers are free to choose from the various products on offer, and they decide, with due cognisance of all of the pertinent factors that constitute 'fitness for purpose' which product best suits there needs. How is a punter selecting an MP3 player from all those on offer different, in principle, from a cheetah selecting prey? In both cases they ultimately lead to extinction of the 'unfit' and survival of the 'fittest'.

 

[Southwind17]

Southwind, my point with the ladder is not that an evolutionary algorithm couldn't get the same result as design (it could), but that the actual process of intelligently devoping the solution to such a trivial problem is not akin to Darwinian evolution at all. It was really answering fishkr's points rather than yours.

Instead of selecting the best examples of random solutions and "breeding" from them, certain solutions are "obvious" and the initial solutions would be developed in the direction of the obviousl solution. Like Lamarck there is "directed change", and no need for selection.

The Newton-Raphson Method of successive approximation is another example I can think of here. Supposing you are trying to find where a funcion is zero, f(x)=0. You have a value at an inital x-value, and know the gradient, so draw a tangent along this line until this line is zero. You then calculate the value of f(x) at this point and repeat untill you get an adequately accurate result.

The process is iterative, but each successive x-value is chosen by drawing tangents to lines and reading their intercepts with the zer-line. There is no "variation" in the choice of subsequent x-values with such a process, beyond the accuracy of the calculation. Of course this falls down for nonlinear systems....

Am I correct in saying your view is that "intelligence" is merely a shortcut?

That's exactly what I'm saying jimbob. In fact, I've explicitly said this on numerous occasions.

 

[Southwind17]

No you can't. You cannot show that the majority of technological development is like your extended analogy because of the simple fact that it isn't.

Whoever said anything about the majority of technological development being like the extended analogy?

Some statues of people are made of marble.
That doesn't means that describing marble as looking like people is at all helpful.

I agree. Did I say it is?

You take a category, define an extremely rare hypothetical subset of that category, then try to pass that off as representative of the category.

Hypotheses, by definition, are extremely rare. In fact, I'd say they're unique. They're used to prove a point though, which mine do.

It simply does not work as an analogy to technological development.

It very simply does.

If you actually created your automaton scenario it would go beyond an excellent analogy and into the category of an excellent model. Unfortunately none of that helps the OP as that model would still not be representative of technological development.

So forget about technological development and go with the automaton scenario instead. It still serves the purpose of the OP, but in a different way.

 

[Southwind17]

Southwind, do you accept that even wioth your "let the maket decide" evolutionary algorithm story, each Successfull variant is chosen by an individual customer, who is an intelligent agent.

Yes, I do, but I fail to see the significance of 'intelligent agent'. Is a cheetah not an 'intelligent agent'? As I wrote above, what's the difference, in principle, between a punter assessing the pros and cons of one consumer product over another and a cheetah assessing the 'pros and cons' of one particular animal for dinner over another?

The selection is thus intwo parts,: Firstly by many different intelligent agents, and then according to other selection criteria to cull out the failures. In any system with finite resources the second is also vital. Every realistic system as finite resources.

Can you please clarify the first part. Are you alluding to consumer product designers/manufacturers or consumers? Also, I'm not sure why you're emphasising the finiteness of resources as important. Can you expand on this too please.

 

[amb]

So what is the only single point of correspondence that you believe technological development displays (incremental change over time, presumably), and how many other potential points of reference do you consider there are (natural selection, i.e. survival of the fittest, presumably)? That makes two.

So why do you consider the retail environment as NOT analogous to the 'biological environment' where natural selection occurs? Consumers are free to choose from the various products on offer, and they decide, with due cognisance of all of the pertinent factors that constitute 'fitness for purpose' which product best suits there needs. How is a punter selecting an MP3 player from all those on offer different, in principle, from a cheetah selecting prey? In both cases they ultimately lead to extinction of the 'unfit' and survival of the 'fittest'.

Survival of the fittest, yes, but not always for the best, Technologically at least. Remember Betacord versus VHS. Betacord was superior, yet VHS survived and prevailed over the better technology. We have a similar situation now with Blueray versus HD players, LCD versus plasma and soon, [I hope] Laser TV.

 

[Southwind17]

Survival of the fittest, yes, but not always for the best, Technologically at least. Remember Betacord versus VHS. Betacord was superior, yet VHS survived and prevailed over the better technology. We have a similar situation now with Blueray versus HD players, LCD versus plasma and soon, [I hope] Laser TV.

I agree. There are other "commercial factors" at play with certain technological developments, but not always.

 

[mijopaalmc]

That's exactly what I'm saying jimbob. In fact, I've explicitly said this on numerous occasions.

But the Newton-Raphson method is a deterministic and systematic process; there is nothing random or haphazard about it. As such, it constitutes and essential difference between what can happen in technological development (sometimes random and haphazard, sometimes deterministic and systematic) and biological evolution (always random and haphazard). If the sometimes random and haphazard, sometimes deterministic and systematic changes in development are replaced with the always random and haphazard changes in evolution, development ceases to be development and becomes evolution.

 

[amb]

The gulf between technological evolution and evolution by natural selection is as wide as
the Earth to the moon. There is no comparison whatsoever. Tech. evolution is guided by man. Evolution of the species is guided by ''natural selection'', survival of the fittest if you prefer. And the twain shall never meet.

 

[Henners]

The gulf between technological evolution and evolution by natural selection is as wide as
the Earth to the moon.

Which is to say: the difference between a whole planet and waving your bare behind.

 

[cyborg]

However all our ancestors did manage to reproduce, which is the only "test". The optimisation is towards "variants" that manage to produce at least one breeding offspring per parent.

Then it's not towards reproductive optimiality is it then?

Without the variants "breeding", nothing.

Jimbob, if "breeding" stops then there are no further iterations are there?

If the selection is completely haphazard, then there will be no selection pressure.

I don't see anyone defining the selection pressures out here, do you?

You leave the system long enough and it'll formulate it's own, natural, selection pressures iff:

1) Not all current objects in consideration in the current iteration will be part of the next iteration
2) The selection of which objects are and are not included is self-defined by the objects themselves

No, you keep asserting that there is no randomness; there is strong evidence that quantum events, like radioactive decay are truly random, these can cause mutations, so at least some mutations are random, even if most might be merely pseudorandom.

I cannot understand that random stream of text.

(You seem to forget that computers are strictly non-random by definition - if you're going to be silly and talk as if the fundamental substrate of reality is ultimatelty consequential in these models then so will I).

At this stage I would be asking how the simulation worked as you seem to be saying that there is no selection criteria.

Predator/prey.

1) Not all current objects in consideration in the current iteration will be part of the next iteration

Predator eats prey - survives.
Predator does not eat prey - dies.
Predator replicates.
Prey eats grass - survives.
Prey does not eat grass - dies.
Prey replicates.

2) The selection of which objects are and are not included is self-defined by the objects themselves

Predator eats prey.
Prey avoids predator.

I imagine at this point you're screaming, "OH THE HUMANITY, WHERE IS THE RANDOMNESS!?!?!"

So you are selecting on "survivability".

Kinda impossible to avoid - objects not utilised in the next iteration are, well, not utilised in the next iteration.

That's true of any algorithm.

Did the surviving code "reproduce". If so then it is self-replicating according to my definition.

See above. I don't think you're considering the pure abstract of what it must mean for code to "reproduce". (Any time data is copied it's "reproduced" - think about the etymology of that word.)

After a certain time, those organisms that are still surviving are copied, but with (pseudorandom?) alterations. This next generation is left in the same fashion and the process repeaded many times.

Don't even need the alterations to prove my earlier points - the newer simulation didn't add a whole lot of difference. Predators/prey (effectively) switching roles doesn't significantly change the dynamics.

If that is the case, you are selecting for organisms that survive and using "artificial selection" to "breed" them.

What exactly is artificial about it? Be consistent now - you said artificial selection is where one tries to attain a specific variant as a goal. I have no specific variants in mind at all. I don't control the simulation once I start it (unless I choose to).

In this case the selection would not be for surving for any length of time, but for reproducing, which would include "surviving long enough to reproduce". It is quite possible that a "parent" organism could evolve to sacrifice some if its resources to boost the reproductive success of its offspring.

I have no idea, none whatsoever, as to why you think what you proposed is significantly more accurate other than for the obvious reason that it is a richer world model.

There's more functional possibilties afforded the variants by the world itself which allow for more reproductive strategies but apart from that I fail to see why what I had made previously is more "artificial" than this.

 

[cyborg]

Am I correct in saying your view is that "intelligence" is merely a shortcut?

If you know the answer then the question is trivial.

How exactly do you solve the ladder problem if you are ignorant of the problem space?

As I've said before it is all very well talking about how "obvious" the solution is from your position of standing on the shoulder's of giants - much of the hard work has been done and you can pluck the predefined solutions from the air and pretend as if they were always "obvious".