Sleepy wrote:QED wrote:How could we not be entitled to declare that we have proved there is a force that can create the "appearance of design" when we can be delivered with "apparently designed" artifacts like those coming out of NASAs evolvable systems group? Naturally it requires a full understanding of the processes involved, but that's something I am only too pleased to discuss.
I haven't studied the field in great detail but I am happy to accept how wonderful such systems work in the, workshop, technological and theoretical world. Can I first ask if these systems exhibit progressive developmental change on the 'stasis' followed by 'sudden change' rules which would allow us to apply such systems to what we see in the fossil record?
I have every confidence that Punctuated Equilibrium would indeed be seen if the environment or selection criteria was to change following a period of stasis. But the typical evolved product is developed at an irregular pace anyway, subject as it is to the internal states of it's algorithm despite the fact that the environment and selection criteria typically remain static. This is because the method predicts fits and starts of developmental activity anyway -- every so often "breakthroughs" are "discovered" by the algorithm that leads to higher efficiencies in meeting the selection criteria. This might have parallels in the transitions between the earliest periods e.g the Ediacaran to the Cambrian where a gearing-up of evolution is seen via the "sudden" abundance of more readily fossilizable organisms.
Sleepy wrote:
Whilst I don't doubt that artificial evolution systems create all sorts of wierd and wonderful things thats hardly evidence in the natural world.
With the greatest of respect your admission is my key concern. You say you have no doubt that there is a
force for creation in the logic "captured" by Genetic Programming. Many "creationists" deny that such a force even exists preferring to think that only Man and God have the capability to design and create complex artifacts. At least we have eliminated this possibility! The question then is one of the likelihood of this being the correct model for the natural version of evolution. I think it would be quite remarkable if the observations of nature (e.g. Darwin's own) were to inspire the development of automated design generators over a hundred years later that
actually work -- only to find that those initial observations and understandings were in fact erroneous.
Sleepy wrote:
You may think this is weakness or lack of insight on my part but I count this as a strength, it prevents me from killing people. I know first hand all too well the effects of models and principles which work wonderfully in theory and often in other disciplines but did not work in the natural world (my patients). I practice evidence based medicine for good reason.
A wise precaution I agree. But I am arguing that we are privy to some very significant evidence in being able to evolve novel (even patentable!) designs using our technology to implement a theoretical methodology draw from our observations of nature.
Sleepy wrote:
Models don't cut it with me my friend. Sorry if this is frustrating to you, you are not going to convince me that because it works on a computer or a workshop it works in all parts of nature without a body of evidence to support it.
Is this wholly reasonably I wonder? It works in nature in so much as the principle is universal. We can't escape it by switching from one environment to another. Without any practical limit to the capacity for complexity in any evolved product we have to have a serious reason to reject it over it being merely a matter of taste. Without even delving into the depths of the process isn't it at all compelling that if something is built to a plan and the plan is passed from generation to generation with minor, random, alterations then the product will adapt and explore the terrain of what is possible and practical? I personally can't imagine not being compelled by this simple process.
Sleepy wrote:
You are arguing micro to macroevolution here and are going to have to do better than modeling it to prove it happens, just as ChosenAndAct is going to have to do better than hoping it works to prove it happens.
But surely it is happening in the evolution of a spacecraft antenna for example? This goes from scratch to a product that meets it's selection criteria without any evident boundary (or anticipated boundary). There is nothing in principle that suggests such a boundary -- other than stepping over a particular philosophical divide.