The analogy requires intelligence, evolution doesn't. Evolutionary algorithms can show how many common aspects of design are not needed to create "designs", but one still needs to remember that natural selection is missing.Oh jimbob, even if you were right, please tell me you have more to offer in your defence of not accepting the analogy than this! Even mijo seems to be coming around to the idea now:In the story #1059, the rule is "take the product to school for x days and try to sell it, if it sellls try to copy it".
This is an externally-applied rule-based system.
There are no rules in natural selection. If an organism fails to reproduce, it doesn't reproduce.
For evolution you need both imperfect copying (mutation) and natural selection.
Without self-replication there is no natural selection. Evooutionary algorithms simply need imperfect copying, and arificial selection. The selection ultimately the result of an intelligent agency. Even, as is plausible, if the rule-based system had itself beendeveloped with an evolutionary algorithm, then the selection criteria for that algorithm would ultimately come from an intelligent agency.
There is no need for an intelligent agency in defining the selection criteria for evooution. Indeed I would argue that, as soon as an intelligent agency starts selection with intent to direct the development of the subsequent generations, then there is artificial selection as well as evolution taking place.
The dolphin and cheetah do not decide to perform selective breeding on their prey; they are part of the myriad selection pressures on the other organisms in their ecosystem, which includes their predators, competitors and prey.That isn't an externally-applied rule, that is just what happens.
"That is just what happens"! So the dolphin and the cheetah 'just happen' to pursue prey until they land lucky? They don't apply some selective criteria in 'deciding' where to hunt, at what time, which prey to pursue, for how long, what dangers to look out for, etc? It's no different from my boys going to school each day: Get up, get dressed, have breakfast, brush teeth, go to school, have lessons, have lunch, have more lessons, come home, etc. THAT IS JUST WHAT HAPPENS, I'm sure they'd have you believe!
Another difference between evolution and technological development is that the selection pressures are far more varied in evolution. This is only a "typical signature" but is worth thinking about.
Further obvious differences
If a design failse because of a particular failure mechanism, then the design is often altered with the express aim of fixing that fault. Features which are considered "good" even of a failing design, could be kkept.
Talk to the story jimbob. Show me where Sam's designs were altered with the aim of fixing the 'fault'. Show me where Sam consciously considered some of the features 'good' in his initial and subsequently failed designs. No doubt in one of his initial failures he happened to have the battery connected to the switch but not the bulb. Show me where he recognized that and consciously retained that arrangement whilst he messed around with the switch configuratioin only. SHOW ME!
The point is not how an unrealistic model of how technological development can be constrained to almost fit into an analogy of evolution, the point is that technological development can produce products that evolution can not.
This being a prime example:
A fluorescent mouse could evolve, given the right conditions, it would be more likely (but still unlikely) with selective breeding. However there is no chance that a fluorescent mouse would evolve with the same gene sequence as a jellyfish.
Intelligent agencies can produce artifacts that evolution can't.
Sam tries to sell his goods. Whilst he is waiting he has not yet sold his goods. How does he decide when to give up on a particular design? Even if he rolls dice to determine the time period, he is still following arbitary rules.This story shows an (ineffective) evolutionary algorithm working. It does not show how technological development works in reality, even when it does use evolutionary algorithms.
Where is Sam's evolutionary algorithm? It does show how technological development works in reality - the story is real - both Ollie and Sam developed their technologies!
The issue isn't with the analogy of random mutation, but with how the selection is performed.
Show me an example of selection which does not require either rules or intelligence, and I will show you an example of self-replication.
I've done exactly that above with the cheetah and dolphin, indeed I believe it shows both rules and intelligence!
The dolphin and cheetah do not decide to perform selective breeding on their prey; they are part of the myriad selection pressures on the other organisms in their ecosystem, which includes their predators, competitors and prey.
No."The market decides" The rule is that one chooses to make copies of those variants that sell within a certain period.
And in nature "the environment decides" the rules that apply to make copies of those variants that survive for a minimum period. No difference jimbob - the market is the environment.
There is no "minimum period" in natural selection. Gnats and elephants have completely different generational lengths, this generational length is also a result of evolution. If an organism reproduces, and makes reproducing copies of itself, then it could be considered to be a reproductive success.
There is no way that anything akin to natural selection, with no rules and no intelligence required, can work without self-replication. If the "replication instructions" are not carried within the varient, then destruction of the variant does not destroy the "instructions". The variant could be recopied again.
Without self-replication there is no way for the variant itself to pass on the "instructions for copying", and there is no need for the variant to survive for these "instructions" to be passed on. The success of the variant has to be assessed according to some arbitary criteria. Arbitary selection criteria require intelligent agencies.
Natural selection is not arbitary, that which replicates, replicates.