cyborg
deus ex machina
- Joined
- Aug 12, 2005
- Messages
- 4,981
Our ancestors were all sufficiently adapted to reproduce. It didn't matter that other organisms reproduced more "optimally", as long as our ancestors reproduces suffficiently well, (whcih they obviously did).
Then we can agree that evolution is not necessarily geared towards optimal reproduction?
By any sensible definition of "optimal", "sufficient" is not optimal.
This is where I dislike articuletts statements about the information that is "best at getting itself copied". If there were many organisms, and one was far better adapted to reproduce, but another orgainsm was neither predator, prey, nor competition for this organism, providing that both have sufficient offspring , both would reproduce and evolve; should neither be sufficiently good at reproducing, it wouldnt matter if one was "better" if better is not good enough.
In a situation with infinite resources competition is avoidable.
As soon as resources became an issue the simple fact is that it is the organism with greater numbers and a greater reproductive rate that has an obvious advantage: more copies around means more chances for those copies to make more copies and they make more copies more quickly.
Imain two populations oone with an average reproductive success of 0.5 reproducing offspring per parent and noe with an average reproductive success of 0.99 per parent, both would become extinct, but one is "better" than the other.
Well yeah - that's kinda how numerical comparisons work.
It's a bit like complaining that 0.49 is less than 0.5 but 0.5 gets rounded up to 1 - that's the consequence of the definitions you used.
You are free to propose any other metric you desire to rank these things.
Why is that not self-replication of some type?
You tell me: you're the one who doesn't like the idea of being able to separate the "self" part from the "replicate" part and yet these abstract entities don't, in any physical sense, "self-replicate".
If the selection is self-defined, then it is self-replication.
That is my point.
It is? How in the hell does that follow from anything you've said previously?
This was a derail, but you keep asserting that there is no true randomness,
No I don't.
I assert that the existence or not of true randomness is irrelevant to the evolutionary principle. It is sufficient that there is variation. True randomness simply has some nice mathematical properties that makes it better than less-than-random mechanisms for these sorts of things. The fact that pseudo-random - i.e. the not random at all type of random - is often good enough for such tasks doesn't seem to register with you for a reason I cannot fathom.
No I am asking why you think that is not self-replication if
2) The selection of which objects are and are not included is self-defined by the objects themselves
I don't see what "self-replication" has to do with that. The statement refers to how a natural selection can form, not how the entities got there in the first place.
Southwind's algorithm isn't. It could just as easily be altered to select for non-selling variants as for sellin variants, especially with the infinite resources allowed.
Uh, then guess which objects are part of the next iteration?
The non-selling items.
Guess which objects are not part of the next iteration?
The selling items.
They are taken out of the shop, never to be known of again whilst those left on the shelf propagate their economic incompetence...
Reproduce itself then.
So you won't consider it self-replication unless the data has the data for how to copy the data?
That seems a little wasteful when I can just centralise the data copying mechanism and achieve exactly the same effect...
Self-selection is the most important feature of self-replication, which I am agguing is needed for evolution.
I think you are getting quite muddled here: how on earth is self-selection a feature of self-replication?
You need a source of variation to work,
What you seem to have failed to grasp is that populations sizes themselves are a source of variation!
If there is not self-selection, which is the important feature of self-replication, then something else performs the selection.
Which is?
In your model, you have said that there is self-selection. Your system sounds as if all variants that are surviving at a certain time are copied imperfectly. Is this the case? In that case that is equivalent to self-replication at a fixed age, and with no ability for reproductive age to evolve.
Changing a constant to a variable is always possible. The point is that doing so here would also not significantly change the dynamics of the situation.
You seem to be obsessed with the overall feature-set of the available phenotype and missing the point about how little needs to actually be variable for the same principles to express themselves.
However this simulation is not in any way akin to technical development.
I think you're at a disconnect here somewhat. First let us be in agreement about the relationship between the evolutionary concept and the simulation and then we can move on to how it relates to more esoteric concepts.
Last edited: