Intelligent Evolution?

[Belz...]

Bloody hell cyborg this link is absolute genius! I think EVERYBODY participating in this thread (excluding Cuddles!) should be REQUIRED to watch the video before they are even allowed to continue in the debate. One would be forgiven for concluding that the guy who posted it based it on my Sam & Ollie story (I promise it wasn't the other way around!). The analogy is a perfect match, in principle.

In fact, I think I'll refuse to debate further with anybody who fails to admit to watching the video, and any further arguments that are posted and that the video adequately disproves will simply be referred to the video! I can tell you now, anybody who fails to watch it will retire from this thread through boredom of receiving my "Watch the Post #1556 link!" stock responses!

Shame about some of the slide transition speeds mind. Just read quickly and/or watch repeatedly is my humble advice!

Fantastic cyborg

And, by the way Cyborg, you might want to show Kleinman that link. I'd like to see what he makes of it.

 

[cyborg]

Ya, really.

Then please give some examples of modern inventions that fit into this categorisation!

I was discussing how Southwinds 'evolution' robot won't design things that resemble our current technologies.

And it won't because you have an odd idea of how it is supposed to work.

Does biological evolution 'ignore the physical dissimilarities of the system to note that which is similar'?

No evolutionary process does that - the point is that doing so is not necessary in order for common "design" artefacts to be observed. As I pointed out earlier there are plenty of examples of "good" design in biology as "bad" design from a human perspective. There is no such bias towards one or the other so why exactly should we expect to observe anything but what we observe currently?

Remember that the biological evolutionary process is restricted to use artefacts within the biological design space. It could not evolve aeroplanes made of metal - it had to make birds.

Its products are not - they exist in a wider "world". This wider world feeds back into the evolutionary system. They are inexorably entangled - one driving the other.

So what is it that "intelligent design" can produce that "evolution" alone cannot? All products in a product world are possible in both.

So what is it that "intelligent design" can do that "evolution" alone cannot? It can bridge product worlds.

 

[cyborg]

So if you have enough 'offspring' in each generation, and combine with the enumerable 'offspring' of many other components, one of them will work. Is that a good comparison to biological evolution? In each generation are 30 different legs, connected 30 different torsos, 30 different heads and so on in order to come up with the next generation human?

You didn't watch the Blind Watchmaker video did you?

Human components, and components of components are not evolving seperately.

What is the vast difficulty you people have with traversing model levels?

Consider simpler, molecular sized components. Think "evolution of multicellular organisms".

 

[jimbob]

Remember I am not talking about the modified analogy, but the general field of technological development.

Ignoring the discussion about the OP for a while, and talking about generalities:

Both are iteritive processes, and thus both show "improvement" over time.

However in evolution, a failure to reproduce is negative information (there isn't really even the instruction "don't try that again". It is just that successes are all that are copied).

In engineering, people often make test structures that are designed to test particular aspects of their designs, these failures are then analysed, and then future designs are altered in light of the information received from these experiments.

Also, if an engineering artifact fails in use, often (usually), this failure is analysed to determine why it failed, so that remedial action can be taken.

This is not analogous to any process in evolution.

I genuinely don't understand you, Jim. Haven't you read the whole thread ? If you did, how can you continue to use this line of argument ? Haven't we established that intelligence is irrelevant to the analogy ?

No, because without self-replication, the selection process has to be somehow defined. The copiers have to be given instructions on what to copy. Letting the market "decide", whilst comming close, still doesn't fully answer how exactly this would work. I am arguing that this is because either self-replication or intent is actually needed.

In classical engineering, without evolutionary algorithms, a lot of engineering effort is devoted to understanding what might go wrong, what has gone wrong, and how to prevent things going wrong. This involves analysing failures.

If you are arguing that evolutionary approaches could in principle replace such effort, and produce parts which meet the specifications, then I would agree with you.

If you are arguing that because evolutionary approaches could produce equivalent structures, the fact that the system involved intelligence is irrelelevant, then I would disagree.

If you consider the results of individual iterations, or even between individual releases, then the evolutionary approach would show a different signature over the history of the different releases.

The (humanly) "intelligently designed" system is unlikely to have unimportant "neutral" alterations between releases, whilst the result of the evolutionary algorithm is. As well as this I would argue that information obtained for the classical approach has not undergone any process remotely analogous to mutation.

Which statements below do you disagree with?

1) Evolution requires the equivalents of both variation and natural selection.

2) Variation arises in any process where imperfect copies are made.

3) Natural selection arises in any process where (perfect or imperfect) self-replication occurs.

4) Evolutionary algorithms, because the products are not self-replicating, require some imposed selection process, either intelligent choice or algorithmic.

5) However, they can demonstrate the power of evolutionary approaches.​

 

[jimbob]

Originally Posted by cyborg
Can "develop an NOR gate" be a natural goal?



The intelligence is illusionary

Exactly-- like a spider "must build web that successfully catches prey" is a natural goal--unstated and without intent... but it must work or the information to build that spider and her webmaking prowess doesn't get passed on. A web looks designed... So long as you have input, replicators, and a selection process you can't help but generate the appearance of design. Whether your goal is to "solve a problem" or do what is programmed via your genes-- the results determine whether the information produced is replicated into the future. Successful information goes on after you die. So does other kinds of information with a trick-- viruses, religions, superstition... we copy those things too... in the process of doing what we've been "programmed" to do.

No, the "goal" is to make replicaors; a NOR gate or a spider's web are means.

This is why I don't like questions like: "what are lungs for"
I would prefer: "What benefit does the posession of lungs confer"

It is more ugly, but avoids the pitfall of thinking that they have a "purpose". They do something that happens to help the organism reproduce.

 

[cyborg]

I am arguing that this is because either self-replication or intent is actually needed.

If I thought you had the imagination to see how it works I would ask you to consider the factory building parts to produce a new factory. I would then point out this is basically how cells "self-replicate".

I would but I know you will say this is not valid.

If you consider the results of individual iterations, or even between individual releases, then the evolutionary approach would show a different signature over the history of the different releases.

And what do you conclude when that signature is not observable jimbob?

Because it seems to me that you would be forced to conclude, with the IDers, that in lieu of this information then a "designer" is absolutely responsible.

This is because both of you still fail to comprehend what design must be contingent on to work.

So I'm going to have to ask you jimbob, in order to perform the critical "intelligent" analysis in a non-evolutionary way how do you:

1) Determine "neutral" design features
2) Determine "good" design features
3) Determine "bad" design features

?

 

[cyborg]

No, the "goal" is to make replicaors; a NOR gate or a spider's web are means.

Arbitrary distinction.

The "means" to replication are the "goals" of replicators.

They cannot simply "replicate" just like that. There are rules.

 

[Belz...]

No, because without self-replication, the selection process has to be somehow defined.

Except information doesn't self-replicate, either. And that's what we're talking about, here. At least some of us.

1) Evolution requires the equivalents of both variation and natural selection.

Yah.

2) Variation arises in any process where imperfect copies are made.

You mean, in the context of biological evolution, or in the general sense ? Biologically, sure. But the source of the variation is irrelevant.

3) Natural selection arises in any process where (perfect or imperfect) self-replication occurs.

False. Replication itself doesn't produce selection. You need other parameters, as well.

4) Evolutionary algorithms, because the products are not self-replicating, require some imposed selection process, either intelligent choice or algorithmic.

Huh ?

5) However, they can demonstrate the power of evolutionary approaches.

Okay, you lost me there.

 

[Walter Wayne]

Then please give some examples of modern inventions that fit into this categorisation!

Did you read the rest of the post.

You didn't watch the Blind Watchmaker video did you?

Did you read the post where I mentioned several characteristics of the video?

I see no point in conversing with someone so disinterested in an exchange.

Walt

 

[cyborg]

Did you read the rest of the post.

Yes.

Did you read the post where I mentioned several characteristics of the video?

I had to search to find which post you mean... I fail to see how any of the characteristics you addressed are relevant since they were mainly along the lines of, "if things were set up this way then it might not produce the product you want."

Yes. I know that one. That's the anthropic argument.

Now: tell me why it cannot produce products of the same type not of the exact same configuration?

I see no point in conversing with someone so disinterested in an exchange.

I've asked you for examples. You didn't provide them.

Where's the disinterest in exchange?

 

[articulett]

And, by the way Cyborg, you might want to show Kleinman that link. I'd like to see what he makes of it.

I did a while back. And I linked a math model that used what we know about how genomes actually evolve to show how readily what we observe can come about. But he's still plodding along using his formula to illustrate that point mutations couldn't have caused all the genomes we see. (Perhaps not... but that's not how it happened anyhow.)

They are impervious. They cannot take in new info. They seem to only be able to focus on their single point until they've convinced themselves they won. It is maddening. You feel like your are drawn into a game where your attention keeps their delusions about themselves and their expertise alive.

 

[President Bush]

You seem to be suggesting your mind is running an evolutionary algorithm.

Why, yes. It is.

Wouldn't that mean your mind is something which was designed?

I think the problem here is that everyday terms like "design" and "intent" are being used - by analogists - as if they were terms in some meta-language... out to construct a definition of truth for everyday language

If semantic terms such as these are to be introduced into an understandable meta-language they need to be to be introduced into the meta-language by definition.

 

[Belz...]

I did a while back. And I linked a math model that used what we know about how genomes actually evolve to show how readily what we observe can come about. But he's still plodding along using his formula to illustrate that point mutations couldn't have caused all the genomes we see

Kleinman's using formulae, now ?

 

[Belz...]

Wouldn't that mean your mind is something which was designed?

I think the problem here is that everyday terms like "design" and "intent" are being used - by analogists - as if they were terms in some meta-language... out to construct a definition of truth for everyday language

If semantic terms such as these are to be introduced into an understandable meta-language they need to be to be introduced into the meta-language by definition.

I see you decided to go on a tangent.

What I meant is that the brain works somewhat in the same way, information-wise. New information gets added (at "random", in the sense that we can't tell what this new data will be), and very often selected for through time. The ideas and thoughts that offer an "advantage" as far as the mind is concerned, are conserved. Heck, they're even "passed on" to other people whenever we can.

 

[articulett]

Kleinman's using formulae, now ?

Oh, I meant his EV point mutation model... that he keeps trying to get Paul to generate formula for...

Or maybe I meant formula-- like the stuff they feed babies... I don't know.

I avoid that thread... it's been more than a year of the same, hasn't it?

 

[articulett]

I see you decided to go on a tangent.

What I meant is that the brain works somewhat in the same way, information-wise. New information gets added (at "random", in the sense that we can't tell what this new data will be), and very often selected for through time. The ideas and thoughts that offer an "advantage" as far as the mind is concerned, are conserved. Heck, they're even "passed on" to other people whenever we can.

He's very tangential... I think he might be a chatbot. Your explanation is fine... he won't be able to comprehend it. None of the "self appointed experts" will. The use of pedantry, sarcasm and generally cluelessness always reminds me of Kleinman-the-impenetrable.

And they're never even witty though they seem to imagine themselves so--but like their expertise, it only exists in their head. I love the insertion of smilies that they think enhance their cleverness (reminds me of the departed hammegk). Soon I imagine they'll be engaging in large and colorful font abuse to slam their "pointless points" home.

I have the whole lot of them on ignore--so some parts of this thread is just "stripes" of them-- I figure if they say anything noteworthy, the intelligent folk will comment. I imagine they can't converse with each other, because each can't understand the other and each is sure they are the ones that truly make the most sense. It's a weird phenomena to observe. They're all conversational mosquitoes imagining themselves a major part of the conversation making major points-- and no one else knows what the heck they are trying to convey-- because the only thing that comes across is the blinking "I'm clueless" aura.

 

[Belz...]

Oh, I meant his EV point mutation model... that he keeps trying to get Paul to generate formula for...

Or maybe I meant formula-- like the stuff they feed babies... I don't know.

I avoid that thread... it's been more than a year of the same, hasn't it?

Yes. Dr Adequate is now of the opinion that Kleinman is mentally ill. I'm beginning to agree.

 

[jimbob]

2) Variation arises in any process where imperfect copies are made.

You mean, in the context of biological evolution, or in the general sense ? Biologically, sure. But the source of the variation is irrelevant.

I am talking about any system with imperfect replication. This does include (living) organisms.

If the "variation" arises as a result of a deliberate process, in technological development, where someone has produced alterations in response to previous failures, then the type of "variation" is different to the type of variation arising from mutation. Surely describing a process such as this as "evolution" is unhelpful; even if you like the idea of "memetic" evolution.

There is another point.

Top-down design can only work for systems that the "designer" can understand, these have to be less complex than the "designer". Selection (artificial or natural) of systems that have random variations are not constrained in this way, because there is no need for understanding.

ETA: For top-down design to work, the important* functioning of the system needs to be understood, not the entire system

*i.e. what the designer wants to happen and why.

I can design a chair because I can understand the important aspects of the system's function (needing to hold up a person's weight and fit them comfortably...)

Random mutation provides both "improvements and degredations" in the information in organisms, natural selection culls out the degredations, and keeps the "improvements". It is a two-stage ratchett.

3) Natural selection arises in any process where (perfect or imperfect) self-replication occurs.

False. Replication itself doesn't produce selection. You need other parameters, as well.

Why is that false?

If a system produces many perfect copies of itself, then the population of "replicators" would rise but it would eventually be limited, in a malthusian fashion due to competition for finite resources, or more "active" destruction from the environment.

If there are many slightly different replicators, all producing identical copies, then some "designs" would be better suited to the environment. These "better" variants would have larger populations.

What else, other than (perfect or imperfect) self-replication is needed for natural selection?

Of course, if the self-replication is perfect, then there could be natural selection, but there is no way for new variation to arise within the populations of self-replicators, so there would be no evolution.

If the replication is imperfect, then the variation can be passed to the replicatiors descendents, so there is both natural selection and variation, so evolution would follow.

4) Evolutionary algorithms, because the products are not self-replicating, require some imposed selection process, either intelligent choice or algorithmic.

Huh ?

Firstly, do my clarifications make sense to you? I am arguing that imperfect self-replication leads to evolution.

If there is not self-replication, but imperfect copying by an external system, akin to a magic imperfect-copying machine, then this could not copy everything, so something has to be selected. How does this selection happen? Destruction of the resulting copies would not need to have any effect. It could be copying blueprints. Somewhere, a selection process would need to be used. I am arguing that this process needs to be set-up and this requires intent.

Okay, you lost me there

I didn't mean to put point 5 in there, oops...

 

[cyborg]

If the "variation" arises as a result of a deliberate process, in technological development, where someone has produced alterations in response to previous failures, then the type of "variation" is different to the type of variation arising from mutation. Surely describing a process such as this as "evolution" is unhelpful; even if you like the idea of "memetic" evolution.

Being deliberate is unintentional.

I can design a chair because I can understand the important aspects of the system's function (needing to hold up a person's weight and fit them comfortably...)

And if I do not know how to design to a systemic description or even how to formulate one, then what?

What else, other than (perfect or imperfect) self-replication is needed for natural selection?

The natural.

Somewhere, a selection process would need to be used. I am arguing that this process needs to be set-up and this requires intent.

You're funny jimbob.

Tell me: does a photocopier 'choose' its imperfections in the copying process? Was it an intention of Xerox to build an "almost photocopier" that would select precisely how to not quite copy your document perfectly?

I really am beginning to think there is something broken in your brain that will not allow you to see how you are arbitrarily setting up an "self/non-self" copying dichotomy that does not entail any consequences for the copying result itself.

You had to use another dichotomy to do that: "imperfect/perfect".

But hey, it's not like I'm saying anything new here so why should I really expect you to accept that you're wrong by your own language this time around?

 

[jimbob]

DNA replication is imperfect copying.

Photocopies don't evolve