• Quick note - the problem with Youtube videos not embedding on the forum appears to have been fixed, thanks to ZiprHead. If you do still see problems let me know.

Intelligent Evolution?

The idea that biological evolution and technological development are analogous also showed in What evidence is there for evolution being non-random? posts 1402 to 1598. The basic premise of that thread that, while there were many eminent scientists who said evolution was non-random, there was actually very little evidence that evolution was actually non-random, at least mathematically in the sense "[o]f or relating to a type of circumstance or event that is described by a probability distribution". Even the scientists who vehemently opposed using "random" to describe evolution said that mutation of increased or decreased an individuals probability of survival and reproduction, implicitly connecting evolution to the sense of "random" provided above.
 
The idea that biological evolution and technological development are analogous also showed in What evidence is there for evolution being non-random? posts 1402 to 1598. The basic premise of that thread that, while there were many eminent scientists who said evolution was non-random, there was actually very little evidence that evolution was actually non-random, at least mathematically in the sense "[o]f or relating to a type of circumstance or event that is described by a probability distribution". Even the scientists who vehemently opposed using "random" to describe evolution said that mutation of increased or decreased an individuals probability of survival and reproduction, implicitly connecting evolution to the sense of "random" provided above.

I don't think you understood.

Mutations are random. Their effects are random. Whether or not mutations are passed on is not random. Deleterious mutations are selected against.

Are you familiar with the concept of Hardy-Weinberg Equilibirum? Its stasis depends on randomness in selection. That's its value in understanding what forces drive Evolution. Selection, however, is not random. It is directional, stablizing, disruptive, restrictive, and often surprising, but it is hardly random.
 
I don't think you understood.

Mutations are random. Their effects are random. Whether or not mutations are passed on is not random. Deleterious mutations are selected against.

The point I have tried to make for the last three months has been that the bias in selection does not make it non-random. If two individuals with the same phenotype (individuals with the same phenotype do not have to have the same genotype or genome) have different "fates" (i.e., one survives and reproduces and the other doesn't), selection is random in the sense that there is given probability that an individual of a certain phenotype will survive and reproduce.

Such randomness, though, does not mean that evolution by natural selection cannot occur, as creationists claim it does when they claim that science says that evolution is random. Instead, it means that evolution occurs as a result of convergence of random variables. The repeated sampling that occurs each generation cause a convergence toward an adaptive optimum.

Are you familiar with the concept of Hardy-Weinberg Equilibirum? Its stasis depends on randomness in selection. That's its value in understanding what forces drive Evolution. Selection, however, is not random. It is directional, stablizing, disruptive, restrictive, and often surprising, but it is hardly random.

Hardy-Weinberg equilibrium is in fact not based on "randomness in selection" it is based one equiprobability in selection (i.e., each individual in the population is equally likely to pass on its genes as any other individual in the population.
 
Last edited:
The point I have tried to make for the last three months has been that the bias in selection does not make it non-random. If two individuals with the same phenotype (individuals with the same phenotype do not have to have the same genotype or genome) have different "fates" (i.e., one survives and reproduces and the other doesn't), selection is random in the sense that there is given probability that an individual of a certain phenotype will survive and reproduce.

Such randomness, though, does not mean that evolution by natural selection cannot occur, as creationists claim it does when they claim that science says that evolution is random. Instead, it means that evolution occurs as a result of convergence of random variables. The repeated sampling that occurs each generation cause a convergence toward an adaptive optimum.


1. Biased systems are non-random.

2. Selection imposes a bias.

3. Ergo, Evolution is non-random.
 
1. Biased systems are non-random.

2. Selection imposes a bias.

3. Ergo, Evolution is non-random.

Biased systems are not mathematically non-random; they are still based on probability distributions and are therefore by definition random. In fact, you have insisted that poker is random even though the game is clearly biased towards a player receiving a high-card hand.

Before you start complaining about the definition of "random" that I am using, would you explain how a normal random variable or a chi-square random variable are random yet very clearly non-uniformly distributed (i.e., biased)?
 
Last edited:
Biased systems are not mathematically non-random; they are still based on probability distributions and are therefore by definition random. In fact, you have insisted that poker is random even though the game is clearly biased towards a player receiving a high-card hand.

Before you start complaining about the definition of "random" that I am using, would you explain how a normal random variable or a chi-square random variable are random yet clear non-uniformly distributed (i.e., biased)?





ran·dom (rndm)
adj.
1. Having no specific pattern, purpose, or objective: random movements. See Synonyms at chance.
2. Mathematics & Statistics Of or relating to a type of circumstance or event that is described by a probability distribution.
3. Of or relating to an event in which all outcomes are equally likely, as in the testing of a blood sample for the presence of a substance.

Selection fails to meet these criteria for randomness. The outcome of a slection process is not normally distributed, and therefore fails #2.
 
Selection fails to meet these criteria for randomness. The outcome of a slection process is not normally distributed, and therefore fails #2.

Uh....the normal random variable is not the only non-uniformly distributed random variable. In fact, there are only a few uniformly distributed random variables: discrete uniform, continuous uniform, degenerate, and Dirac.

Could you explain how you came to the conclusion that only normally distributed random variables are random?
 
What about putting this back on t'other thread.

I like "probabilistic" selection. It would probably fit the poisson distribution.

ETA:

The selection pressure is not random.

ETA In a stable environment
 
Last edited:
Uh....the normal random variable is not the only non-uniformly distributed random variable. In fact, there are only a few uniformly distributed random variables: discrete uniform, continuous uniform, degenerate, and Dirac.

Could you explain how you came to the conclusion that only normally distributed random variables are random?

When I took statistics, I was taught that any distribution that was not normal was non-random, and that the analysis showed, within a certain degree of confidence, that the sample deviated from randomness due to some cause.

Now, if you'd like to drag this out, please write out the definition of "random" you believe applies which is actually used in mathematics.
 
Last edited:
Thanks for the information jimbob, which leads me to believe that my analogy does not rely on the invocation (non-religious meaning) of an evolutionary algorithm. I see design with intent and/or forethought simply as circumventing the alternative option in the primary interests of time and cost savings, that alternative, default option being the making of entirely random design changes with no specification and then trialing them 'in the field' (or 'dumb universe', if you like) to see what survives and what doesn't.

I don't really agree, although it is a subtle point.

Most importantly:

that alternative, default option being the making of entirely random design changes with no specification and then trialing them 'in the field' (or 'dumb universe', if you like) to see what survives and what doesn't.

How does one decide whether something has "survived", unless it is either "survived for a length of time, or 'incidents'" (arbitarily defined with intent) or if it has survived to produce as many copies of itself as it can.

The best example for this analogy might be fighter aircraft.

Do you decide to choose those which survive one dogfight, ten , one hundred? Does the machine itself choose? (those hawks which have bred, have bred)...

When you get to computers, what do you choose as the selection criteria?

Design could be argued to be a short-circuit of evolutionary algorithms, but not of evolution.

Even there, going back to the TV with the CRT vs LCD example.

Imagine an evolutionary approach to creating a 14" TV display in the 1980's.

One random variation uses an LCD screen. It is far worse than a CRT display. There is no way that would get selected, unless an intelligence decided that it was worth pursuing because it had potential.
 
Last edited:
When I took statistics, I was taught that any distribution that wasnot normal was non-random, and that the analysis showed, within a certain degree of confidence, that the sample deviated from randomness due to some cause.

Now, if you'd like to drag this out, please write out the definition of "random" you believe applies which is actually used in mathematics.

I already provided a definition of "random" used in mathematics: "[o]f or relating to a type of circumstance or event that is described by a probability distribution". This is the only definition that provides justification for calling the vast array of random variables "random".
 
I will waste no more text on someone who does not read a damn thing I say.

I read them. I understand them. I feel your pain.

Per my sig, the least competent are the most likely to overestimate their knowledge. I think a brief scan of this thread would have most people in agreement as to who such people are. The ones not in agreement, would be those people.
 
I read them. I understand them. I feel your pain.

Per my sig, the least competent are the most likely to overestimate their knowledge. I think a brief scan of this thread would have most people in agreement as to who such people are. The ones not in agreement, would be those people.

I would say that articulett's signature probably describes her the best, ironically.
 
Selection fails to meet these criteria for randomness. The outcome of a selection process is not normally distributed, and therefore fails #2.

He had a huge thread on the topic where many dropped in to say what you said-- but he still thinks it makes sense to sum up evolution as random because anything related to probability in any way is random to Mijo.

Don't even start this--search for threads he started and you'll learn everything you need to know regarding his thinking (or lack thereof) on the subject.
 
He had a huge thread on the topic where many dropped in to say what you said-- but he still thinks it makes sense to sum up evolution as random because anything related to probability in any way is random to Mijo.

Don't even start this--search for threads he started and you'll learn everything you need to know regarding his thinking (or lack thereof) on the subject.

The problem is that neither uniform nor normal distribution is required for something to be mathematically. At its core mathematical randomness is defined on a probability measure. Functions that are defined on a probability measure are random and those are not defined on a probability measure are not random.

What is so hard to understand about that?
 
Per my sig, the least competent are the most likely to overestimate their knowledge. I think a brief scan of this thread would have most people in agreement as to who such people are. The ones not in agreement, would be those people.

I'm not sure it's over-estimation of knowledge articulett, more incorrect preconceptions compounded by dogmatism and possibly plain old misunderstanding, but possibly a combination of all of these. I could be accused of the last, possibly with some justification, which I could humbly and justifiably, I believe, put down to my lack of background in this subject, but certainly not the others. That said, I do not believe that the 'more knowledgable' participants to whom you allude have yet either convincingly countered the analogy or the vast majority if not all of the points that I have subsequently raised in support.

You will see from my OP that I would not, at that time, have been at all surprised to have been immediately shot down in flames with my analogy. The fact that I wasn't, and that both you and Cyborg have argued the case far more clearly and logically than the opposition have, strongly indicates that the analogy has teeth, and probably sharp ones at that.

Having said all that, I don't think I would categorize jimbob as one of those to whom you allude. Whilst he clearly sits in the opposition camp I can at least follow most of what he writes, which seems generally logical, albeit it sometimes flawed. Actually, 'flawed' might be a disingenuous word. I think jimbob is generally only guilty of failing to see my point, but that could also be down to my less-than-perfect ability to argue and debate. I'll happily give jimbob the benefit of any doubt, for the time being, as I reckon he could come around to our way of thinking.

On that note I'll now respond to his last post:

I don't really agree, although it is a subtle point.

Why do you not agree? You sound doubtful of your position - perhaps you could be conviced through further discussion?

Most importantly:

How does one decide whether something has "survived", unless it is either "survived for a length of time, or 'incidents'" (arbitarily defined with intent) or if it has survived to produce as many copies of itself as it can.

The best example for this analogy might be fighter aircraft.

Do you decide to choose those which survive one dogfight, ten , one hundred? Does the machine itself choose? (those hawks which have bred, have bred)...

When you get to computers, what do you choose as the selection criteria?

This is an interesting question, but I suppose you could ask the same of natural evolution. I don't know how many iPods there are now on the planet, but I doubt that any reasonable person would argue that the iPod hasn't 'survived' or isn't a 'successful' design. I suppose the sales statistics form the measure of success in this instance. The Sinclair C5 (remember that one), in contrast, was a 'commercial disaster', to quote Wikipedia, and as a result soon became 'extinct'. Again, I suspect the measure of its failure was its inability to generate sales targets.

In either case, although the sales targets might well be construed as artificial selection, they're not really. Apple or James Sinclair could, at least in theory, simply have developed their respective products and put them on the market with no criteria for 'success'. Whether they became 'successful' or otherwise would, in any event, have become apparent sooner or later by whatever measures would have been deemed appropriate (running out of money to continue production, super-profits, competition?!), and the production would either continue (replication - survival of the fittest) or cease (extinction) (or, of course, evolve to something better!).

To what extent can the dolphin, for example, be considered a survivor? Would you not have considered the dodo a survivor, had you been asking the question, say, 1,000 years before its extinction? 'Survival' seems to me a somewhat arbitrary or relative term when applying it across multiple generations. The Spitfire was a highly successful fighter aircraft, but is now extinct. I would guess it could be concluded that it 'out-survived' its counterparts, but by what measure?

Design could be argued to be a short-circuit of evolutionary algorithms, but not of evolution.

OK, in the context of what's already been argued on this thread by me, cyborg and artuculett, in particular, please offer that argument.

Even there, going back to the TV with the CRT vs LCD example.

Imagine an evolutionary approach to creating a 14" TV display in the 1980's.

One random variation uses an LCD screen. It is far worse than a CRT display. There is no way that would get selected, unless an intelligence decided that it was worth pursuing because it had potential.

I think you might be inadvertently alluding to the widely rejected hypothesis of saltation with this example (incidentally, the piston engine/jet engine analogy counter argument also falls victim to it), or at best confusing two entirely different 'products' with separate evolutionary histories that just happen to perform essentially the same function. The bow and arrow and AK47 might be a good example to make the point, or the cheetah and wolf, if you prefer a biological comparison.

The basic evolution/design analogy need not go so far as the LCD screen to work; it can stop at the CRT and still hold true (a conventional CRT TV could equally have been used for my OP example, but I guess it demonstrates fewer incremental changes compared to the motor car). I think you're still looking for exceptions that 'buck' the analogy rather than examples that follow it, and falling victim to the preconception that the two cannot be likened in the process, as certain other posters have.
 
Last edited:
The fact that I wasn't, and that both you and Cyborg have argued the case far more clearly and logically than the opposition have,

It's pretty simple from my perspective: jimbob, ID and mijo are wasting an inordinate amount of time pissing about whether or not a particular word is correct with no thought whatsoever as to the actual systems being discussed. So we have random, non-random, probability distribution, design, evolution et al banded about as if labelling the thing differently had some sort of consequence to the system.

Not a single one of them has bothered to respond to the algorithmic relations here which cut to the heart of the behaviour of the system. Nor it seems have any of them grasped, despite the fact I have repeatidly demonstrated it, that systems can always be described in terms of their complements.

The Poker analogy: is this a random game? No. It is not. You're not even thinking on the right level if you are asking the question. The question you should be asking is: what are the stragies for winning this game? Now things become interesting.

Formulating a winning strategy for any game is contingent on knowledge about the state of the game as it is, how the state of the game can change and what winning states of the game are. If you have complete knowledge then winning strategies can be entirely deterministic - as was demonstrated recently for Checkers by exhaustive search. The less knowledge about the complete state of the game you have then your ability to formulate 'perfect' strategies is diminished - you are forced, at some point, to consider alternative game states for which you cannot decide which is actually in play. Now if in considering these alternatives what play decision you make is unaffected then this is equivalent to having a deterministic problem. In all other cases there will be an unresolvable choice. This is where statistics come in - in the range of alternatives you then select based on which choice is more likely to form a winning outcome.

Now the reason why it is important to have a properly randomised deck in Poker is that it affords deterministic strategies if it is not since it will be possible to have more knowledge about the state of the game and hence make better decisions. This is why collusion in a Poker game can mean an advantage to the colluders - they have more information. The game, you should note, is the same. Is Poker random? That's the wrong question. The right question is: what is it possible to know and when is it possible to know it?

Asking if Poker is random is idiotic: the rules are set. There is nothing random about it. The question is what the consequences of the rules are. Well the consequences of a system where decisions need to be made on information which cannot be known is that non-determinism will be exhibited. Decisions will literally not be determined based on the thing they are supposed to be.

Now to Roulette. Is it random? Again this is an idiotic question: the rules are set. The consequences of the rules of this game is that there is no information which can be used to formulate a winning strategy that will outperform 'pure' randomness - maximum information entropy.

At this point saying Roulette is basically the same game as Poker because of the involvement of elements which you cannot know looks stupid. Clearly there is a continuum and it's clearly based upon the amount of information you can have about the whole system.

No information -> pure random.
Full information -> pure non-random.
Anything inbetween -> mixed.

If the fallacy of pissing about random/non-random isn't apparent by this point then it's a lost cause. If the fact that an infinitely sized non-random system will pretty much be equivalent to a random one by virtue of the fact you cannot have full information about it doesn't ram it home and the fact that any finitely sized random system will have a full and non-random description doesn't ram it home then we're talking to people who really only can see the words and nothing else.

And what does this have to do with evolution/design? Why is there an analogy? Well if anyone asking this question had actually read what I had said many pages ago they would have been able to see that where ignorance is maximal design is impossible. Where complexity is maximal design is impossible. It is the vast quantities of ignorance that we have about how to design what we want that makes human design evolutionary. Wherever that ignorance is minimal the analogy to evolution looks weakest.

Whatever.
 
And what does this have to do with evolution/design? Why is there an analogy? Well if anyone asking this question had actually read what I had said many pages ago they would have been able to see that where ignorance is maximal design is impossible. Where complexity is maximal design is impossible. It is the vast quantities of ignorance that we have about how to design what we want that makes human design evolutionary. Wherever that ignorance is minimal the analogy to evolution looks weakest.

I'm pleased you finished off with this cyborg. Whilst I understand your statistical explanations I might have been inclined to question their relevance to the analogy, but this is an elegant and fitting epilogue. Well done ;)
 
mijopaalmc; said:
I already provided a definition of "random" used in mathematics: "[o]f or relating to a type of circumstance or event that is described by a probability distribution". This is the only definition that provides justification for calling the vast array of random variables "random".

That's a useless definition! Any physical system has some probability distribution. What is the length of the sleeves of the shirts made in a certain factory in Indonesia? It's a target length, plus or minus a small amount to due the variabilities in the machine, the cloth, the fibers, and other factors. All real systems with real outputs have some sort of probability distribution.

Try again, with a definition of "random" that doesn't include everything in the entire universe.
 
cyborg; said:
It's pretty simple from my perspective: jimbob, ID and mijo are wasting an inordinate amount of time pissing about whether or not a particular word is correct with no thought whatsoever as to the actual systems being discussed. So we have random, non-random, probability distribution, design, evolution et al banded about as if labelling the thing differently had some sort of consequence to the system.

This is a pernicious lie, and you know it. What we are arguing is that the analogy between the development of machines and the Evolution of living things is incredibly weak because design can plan, Evolution cannot, living things have inheritance, machines do not, and a litany of other inherent differences of origin, which have lead to inherent differences in form and function you fail to acknowledge.

Not a single one of them has bothered to respond to the algorithmic relations here which cut to the heart of the behaviour of the system. Nor it seems have any of them grasped, despite the fact I have repeatidly demonstrated it, that systems can always be described in terms of their complements.

I have relentlessly addressed the differences in the algorithms of Evolution and design, outline precisely how they are different.

The Poker analogy: is this a random game? No. It is not. You're not even thinking on the right level if you are asking the question. The question you should be asking is: what are the stragies for winning this game? Now things become interesting.

Your analogy is a tangent.


And what does this have to do with evolution/design? Why is there an analogy? Well if anyone asking this question had actually read what I had said many pages ago they would have been able to see that where ignorance is maximal design is impossible. Where complexity is maximal design is impossible. It is the vast quantities of ignorance that we have about how to design what we want that makes human design evolutionary. Wherever that ignorance is minimal the analogy to evolution looks weakest.

Whatever.

Your analogy is irrelevant to the OP, which is teaching Evolution to a person through analogy to the development of living things.
 
Last edited:

Back
Top Bottom