cyborg
deus ex machina
- Joined
- Aug 12, 2005
- Messages
- 4,981
Non sequitur.
Biological design spaces are a subset of all design spaces.
The difference is irrelevant in the comparison being made.
You will not grasp this.
Intelligent Evolution?
- Thread starter Southwind17
- Start date
Non sequitur.
Biological design spaces are a subset of all design spaces.
The difference is irrelevant in the comparison being made.
You will not grasp this.
articulett
Banned
- Joined
- Jan 18, 2005
- Messages
- 15,404
Jimbob, when you say "the organism copies itself"-- you are invoking a metaphor... and it's somewhat true (akin to saying the sun rises when actually the earth is rotating towards the sun)... but what is really happening is that the information for making a new organism is getting copied and that codes for a new organism. Humans don't divide into two people. All of the copying is done at the "information" level-- (the directions are being copied).... from ERVs to viruses to bacteria to sponges to fungi to us... nothing is actually copying itself. The information is being copied because it has found a vector that will do the job because of something specific about it (the information). The butterfly mutation got itself copied... The butterflies did not copy themselves--they copied the DNA and the DNA copies with the mutation build butterflies that preferentially survived. You are approaching everything backwards. Ideas and designs and phrases and recipes and other "memey" thing get themselves copied similarly...
But I agree with Cyborg. This will not compute. You'll be off on luminescent tangent land.
But I agree with Cyborg. This will not compute. You'll be off on luminescent tangent land.
Last edited:
President Bush
Banned
- Joined
- Mar 27, 2005
- Messages
- 1,506
OK.
I was under the impression that abstraction in mathematics involved abstracting only mathematical concepts (to generalize structures), not the abstraction of (already) abstract ideas or mental symbols into mathematical formalisms.
I humbly confess that I just plain lack the know-how needed to accomplish such a thing.
To allow me (and others) the chance to reason more precisely here, cyborg... would you please be kind enough to summarize each particular concept involved in the analogy of Southwind17 as its (respective) mathematical formalism?
articulett
Banned
- Joined
- Jan 18, 2005
- Messages
- 15,404
cyborg
deus ex machina
- Joined
- Aug 12, 2005
- Messages
- 4,981
I will not speak to Southwind17's analogy specifically on a blow-by-blow basis because this is NOT about starting off from some static point - this is a dynamic discussion and I will therefore summarise my understanding of this issue.
A basic evolutionary step can be formulated as:
G -R-> O:{ G', G'', ... } -S-> O' { ... }: member of O
Data G is modified by a reproduction process R to produce a set of offspring O which is then thinned by selection process S to produce a new set of offspring O' which are designated as 'survivors'. The next 'generation' of entities can be produced simply by repeating this process for every surviving member.
Now we can talk about the consequences of the structures afforded G, the reproduction mechanism R and the selection mechanism S.
Keeping it abstract let us consider data G as a simple number. S will therefore simply select survivors based on a number - we could define a set Gs of G that will survive (using some algorithm to define it perhaps). We can define R as a functional mapping that simply takes one number and returns a set of new numbers.
For example (in pseudo-code notation):
0 -{+1, +2 }-> {1, 2} -(%2 == 0)-> {2}
Or "take G, produce the set O by adding 1 and adding 2 then select the even members of this set".
If R is an identity function (i.e. G -> {G}) then we will see no interesting inter-generational changes other than a pruning of the set O until only those in the set Gs exist.
If R returns a 'random' set O then we will see no particular pattern arising inter-generationally - again we will simply see every member in O that exists in Gs for each generation. G is not related to O - the output is 'non-deterministic'. The only thing of note here is that if the set O doesn't have any member in Gs then the process will reach a state of 'extinction' - which is pretty boring.
Now if R returns a 'non-random' set O then we can say that the relationship between G and O is 'deterministic'. It is at this point that we will also start to see inter-generational relationships arising since there will be a chain G -> O -> O' -> O'' -> O''' etc... where each member of the chain was 'determined' by the last. Carrying on from the last example:
0 -{+1, +2 }-> {1, 2} -(%2 == 0)-> {2}
2 -R-> {3, 4} -S-> {4}
4 -R-> {4, 5} -S-> {6}
...and so on.
Now if R is a function such that it affords the possibility that each possible G can be constructed then we can know that every possible G in Gs can be reached. If not then there may be G that cannot ever be constructed in the set Gs from our starting point and with our function R. So if we modified R such that it was {+1} then we can see that we will never reach any even G - and we'd have immediate extinction to boot. Of course if it was {+2} then the case would be the same except we'd be able to hit a member of Gs each time - but 'knowing' this would require more thorough analysis of the properties of the system.
Now consider the function R of the form {+(0:1)}. Here we add either 0 or 1 but we do not specify which - the choice is arbitrary. Here we have a random change. (And as I showed above putting randomness in the wrong place doesn't produce anything of note so it is important to note this). Here we are guaranteed to hit all members of Gs (providing extinction does not occur) but, unlike the last case, we cannot be sure about how long it will take to do so. In the last case it would take n generations to hit every member of the even set of Gs with n members. Here we cannot be certain about the time.
(And as an aside if we have a sufficiently complex deterministic hidden function that is selecting 0 or 1 'behind the scenes' we can get much the same effect - we are not 'sure' about whether we are adding 0 or 1 this generation).
Now I think I will stop there before relating numbers to physical systems because I suspect the above will cause enough problems as far as gaining consensus between the group.
Last edited:
mijopaalmc
Philosopher
- Joined
- Mar 10, 2007
- Messages
- 7,172
This is unnecessary multiplication of entities to maintain a deterministic universe. We know that certain random systems have to converge on an expected value if the sampling criteria are kept constant over long periods of time.
Why this insistence on hidden variables?
Aren't hidden variable by their very definition unfalsifiable phenomena?
How would one prove that hidden variable exist without in some way proving what they are?
cyborg
deus ex machina
- Joined
- Aug 12, 2005
- Messages
- 4,981
Non-****ing-sequitur. Try to keep on track. Anyway.
*SIGH*
I am not the one proposing ANY physical consequences - I am merely pointing out the mathematical consequences. And that there certainly could be entirely deterministic phenomena that are unfalsifiable by virtue of 'hidden' variables - and I say 'hidden' because they could be in plain sight but just far too hard to analyse by virtue of the complexity of the system - is, I'm afraid, a consequence of the tools we have to reason about the universe. There is, by definition, no way to know.
I am, therefore, flexible when choosing my abstractions. Something you are clearly incapable of doing when you insist, quite irrationally, upon one or the other.
Last edited:
mijopaalmc
Philosopher
- Joined
- Mar 10, 2007
- Messages
- 7,172
Why is that not important?
It allows natural selection.
If the instructions are different, then the resulting "vehicle" is different. If the vehicle fails to reproduce, then the instructions embodied in the vehicle are not selected.
mijopaalmc
Philosopher
- Joined
- Mar 10, 2007
- Messages
- 7,172
cyborg-
jimbob's argument isn't flawed because he doesn't understand what an abstraction is. He's told you repeatedly that your abstraction just doesn't work if you hope to counter intelligent design with your analogy. That said, you are right that biological evolution and technological development are both examples of the abstract concept of "change over time with retention of what 'what works'", but that doesn't provide a full description each process, a task which requires acknowledgment of the differences between between them that you think you have succeeded in "abstracting away".
jimbob's argument isn't flawed because he doesn't understand what an abstraction is. He's told you repeatedly that your abstraction just doesn't work if you hope to counter intelligent design with your analogy. That said, you are right that biological evolution and technological development are both examples of the abstract concept of "change over time with retention of what 'what works'", but that doesn't provide a full description each process, a task which requires acknowledgment of the differences between between them that you think you have succeeded in "abstracting away".
Southwind, the only "intellignt design" that I belive in is that performed by humans. In engineering, for example. Maybe you do not consider that intelligent design.
I regard this as "no chance". Even over the age of the universe.
Could you explain your "spineless" comment?
Do you agree that evolution requires "selection" and "mutation"?
Without self-replication, a copy could be made of a "design" even if the physical structure has been destroyed, because the "vehicle" is divorced from the "instructions". The "instructions" can make any number of copies without any resulting copies being present. A car assembly line does not need cars to make more copies of the cars.
If the car design is to evolve without intelligent direction, then a system of selection needs to be implimented (without selection there is no evolution).
It is possible to define a selection of specifications that the product should meet. It is possible to have technological development as in the stories described by southwind. However in all these cases, there are implicit or explicit selection criteria, and selection algorithm algorithms, which have been chosen by intelligent agencies.
Without self-replication, an external set of selection criteria need to be applied, these can not be instigated, except as a result of intelligent action, nor could they be implimented, except as a result of intelligent action.
As you seem to disagree with this, please give me a convincing example.
If you have a process that makes imperfect copies of a structure or device, but these devices do not themselves self-replicate, then any selection needs to be applied by an intelligent agency.
The lack of an intelligent agency is one of the crucial features of evolution.
Technological development, for the reasons above, requires intelligent agencies.
The analogy in the OP is bad because, although the "variation" part can be made to work, the type of selection described in the OP is inherently dependent on an intelligent agency.
Natural selection is not dependent on an intelligent agency. This lack of dependence is due to self-replication. The variation is due to the impefect copying. Thus imperfect self-replication is required for evolution.
In its original form the OP did not distinguish at all between intentional alteration of a design to fix a certain problem, and random changes as in evolutionary algorithms. Indeed, it seemed to be closer to lamarck's idea that alterations occur in response to needs.
In its original form the analogy described in the OP was indistinguishable from many descriptions by avowed ID proponents.
The analogy has improved, as now the stories do describe evolutionary algorithms, so a small subset of technological development can be considered to be a good analogy for the variation within evolution, but it falls down when we come to the selection part.
A simple addendum, to the stories, that selection doesn't work quite like that, with a treatment of the effects of self-replication would mean that the analogy was no longer so misleading, yet superficially attractive.
It is not an obsession with jellyfish, it is pointing out that whilst it is is conceivable that a mouse could have evolved the same 700+ letter gene sequence as one that is already observed in jellyfish, you would have better odds of winning the national lottery jackpot every week for a year, when entering only one ticket per week (which is 1 in 3.969x10371). In fact the odds are tens of billions of times better by a conservative estimate).
I regard this as "no chance". Even over the age of the universe.
Could you explain your "spineless" comment?
I have previously discussed this elsewhere.
Do you agree that evolution requires "selection" and "mutation"?
Without self-replication, a copy could be made of a "design" even if the physical structure has been destroyed, because the "vehicle" is divorced from the "instructions". The "instructions" can make any number of copies without any resulting copies being present. A car assembly line does not need cars to make more copies of the cars.
If the car design is to evolve without intelligent direction, then a system of selection needs to be implimented (without selection there is no evolution).
It is possible to define a selection of specifications that the product should meet. It is possible to have technological development as in the stories described by southwind. However in all these cases, there are implicit or explicit selection criteria, and selection algorithm algorithms, which have been chosen by intelligent agencies.
Without self-replication, an external set of selection criteria need to be applied, these can not be instigated, except as a result of intelligent action, nor could they be implimented, except as a result of intelligent action.
As you seem to disagree with this, please give me a convincing example.
If you have a process that makes imperfect copies of a structure or device, but these devices do not themselves self-replicate, then any selection needs to be applied by an intelligent agency.
The lack of an intelligent agency is one of the crucial features of evolution.
Technological development, for the reasons above, requires intelligent agencies.
The analogy in the OP is bad because, although the "variation" part can be made to work, the type of selection described in the OP is inherently dependent on an intelligent agency.
Natural selection is not dependent on an intelligent agency. This lack of dependence is due to self-replication. The variation is due to the impefect copying. Thus imperfect self-replication is required for evolution.
I understand how far this analogy goes.
In its original form the OP did not distinguish at all between intentional alteration of a design to fix a certain problem, and random changes as in evolutionary algorithms. Indeed, it seemed to be closer to lamarck's idea that alterations occur in response to needs.
In its original form the analogy described in the OP was indistinguishable from many descriptions by avowed ID proponents.
The analogy has improved, as now the stories do describe evolutionary algorithms, so a small subset of technological development can be considered to be a good analogy for the variation within evolution, but it falls down when we come to the selection part.
A simple addendum, to the stories, that selection doesn't work quite like that, with a treatment of the effects of self-replication would mean that the analogy was no longer so misleading, yet superficially attractive.
cyborg
deus ex machina
- Joined
- Aug 12, 2005
- Messages
- 4,981
Counter intelligent design? mijo - intelligent design doesn't exist in the world I'm talking about because "intelligence" and "design" aren't special entities - they're properties of a more basic underlying system.
FLIP mjio, FLIP.
It's almost like it flies right over your head every time I point this out:
ASBTRACTIONS IGNORE DIFFERENCES.
By their very nature they will NOT be "full descriptions".
FLIP.
No. Organisms reproduce, self-replication is a defining aspect of life.
Planes don't
I fail to see how pointing the vital role that sexual- or asexual- reproduction plays in evolution is somehow "creationist". These are both imperfect self-replication. At the DNA level.
The LMB is part of the MRC, this is an accademically recognised organisation, the "Protein and Nucleic Acid Chemistry" Division is well within the biological science mainstreem.
The statement
Is thus uncontentious and representitive of the consensus viewpoint.
What is your intent?
You asked who the "mysterious mathematician" was, I answered the questionI was agian showing that if something has made it onto a university course, then it is scientifically uncontentious.
The quoted text showed why self-replication is required.
If the odds are so great that it is vanishingly unlikely within the age of the universe, indeed within the best projections for the total lifetime of the universe, then I would say that it is reasonable to say "never".
What do you mean?
Are you denying that the genetic modification was performed by intelligent agents? Or are you claiming that the mouse acquired this gene sequence through evolution?
I am pointing out that there are differences betweeen the results of intelligent action and of evolution. These diffences are because the processes are fundamentally different.
mijopaalmc
Philosopher
- Joined
- Mar 10, 2007
- Messages
- 7,172
Wow, someone is in denial. Intelligent design may not have any scientific backing whatsoever, but that doesn't stop it from being a tremendous cultural force that is competing with evolution by natural for dominance. Or have you forgotten that parents had to go to court to keep intelligent design out of the science class room?
I am not ignoring that abstractions ignore the differences; I am saying that it is not helpful to ignore the differences when you are trying to teach evolution by natural selection or engineering. After all, you would try to teach geometry while ignoring that different geometries result from different parallel postulates.
cyborg
deus ex machina
- Joined
- Aug 12, 2005
- Messages
- 4,981
And some people seem to be making gross presumptions.
You make this sound like it's something new mijo.
I see.
So the argument you guys have got basically doesn't come down to addressing anything I'm saying but basically going:
"Won't somebody please think of the children?!?"
Would it be too much to, you know, actually address the merit of the thing itself rather than the cultural impact it might have?
You are not even in the same universe of discussion as me.
You don't understand abstractions.
cyborg
deus ex machina
- Joined
- Aug 12, 2005
- Messages
- 4,981
The screw in my door is not the screw in my window.
THE IMPORTANCE OF THIS DIFFERENCE DEPENDS ON THE CONTEXT OF THE DISCOURSE.
No.
No.
articulett
Banned
- Joined
- Jan 18, 2005
- Messages
- 15,404
Organisms do not "self reproduce".
As long as you keep missing this, you sound like a bozoface. How do organisms reproduce? They do it by getting parts or all of their DNA into new replicators. Viruses do this and computer viruses do this. Self reproduction ONLY REFERS TO THE INFORMATIONS ABILITY TO GET ITSELF COPIED. The replicator competes in the environment... but the replicator doesn't replicate itself... it copies information which build new things that may or may not be copied.
INFORMATION... JIMBOB.
Until you separate the information from what it codes form you will not be able to have an intelligent discussion on the topic. You can't understand the analogy because of this continual confusion on your part. The only that replicates is the information--that information builds new replicators or new things that can be replicated... but those things don't replicate... only the information to build such things does. Damn. It's like you are playing checkers when everyone else is playing chess. Evolve.
As long as you keep missing this, you sound like a bozoface. How do organisms reproduce? They do it by getting parts or all of their DNA into new replicators. Viruses do this and computer viruses do this. Self reproduction ONLY REFERS TO THE INFORMATIONS ABILITY TO GET ITSELF COPIED. The replicator competes in the environment... but the replicator doesn't replicate itself... it copies information which build new things that may or may not be copied.
INFORMATION... JIMBOB.
Until you separate the information from what it codes form you will not be able to have an intelligent discussion on the topic. You can't understand the analogy because of this continual confusion on your part. The only that replicates is the information--that information builds new replicators or new things that can be replicated... but those things don't replicate... only the information to build such things does. Damn. It's like you are playing checkers when everyone else is playing chess. Evolve.