Your post for debunking Kurzweil doesn't really address what I was talking about. I made the statement that real AI isn't really possible because the debug and error checking system of the AI would be entirely human designed, and this component of the system would define just about everything that the AI was doing at any given time.
Yes it does. Read it again more carefully. I already encompassed the generative essence of that.
http://unheresy.com/Information%20Is%20Alive.html#Algorithm_!=_Entropy... or it's going to mirror biological evolution.
You are repeating what I wrote in my blog before you did.
Since the portion of the human genome pertaining to the brain has an entropy in the millions or billions, each human brain is potentially at least one-in-a-million or one-in-a-billion unique. Notwithstanding that uniqueness, if human evolution was entirely encoded in a finite genome, then it would be mathematically possible for a plurality of humans to have identical brains at some point in time as the brain forms before differentiation from non-identical learning environments. However, the brain is learning and exposed to the environment as it is forming in the womb, thus there is never a point in time where the brain was entirely structured from only the information in the DNA.
Thus evolution is not just an encoding from the environment to the genome, rather a continuous interaction between the ongoing environment and the genome. Thus for computers to obtain the same entropy of the collective human brainpower, they would need to be human reproducing, contributing to genome and interacting with the environment in the ways humans do. Even if computers could do this, the technological singularity would not occur, because the computers would be equivalent to adding more humans to the population.
Haven't read it before. The whole argument basically comes down to, is it possible for machines reproducing asexually to compete with biological genetic diversity. It's a hard call to make with the low overhead of machines creating new machines if they were also competing with each other for resources. If the programs just existed in a vacuum and didn't have to compete with each other or biological entities for resources, then of course they're not going to develop into anything.
Instead of trying to create AI from scratch, with human based error checking and debug rules encompassing all of it's functionality, if all you did was try to digitize a rat brain, the low overhead of machine reproduction could accelerate natural selection so fast that it turns from rat to god overnight, possibly while just sitting inside of a simulator fighting other rats. So then the question is, what is the lowest level organism needed to be digitized to accomplish such a task.
In this model, you're not actually trying to create high level organisms, you're just trying to lower the overhead of natural selection on more primitive organisms. If the machines only used asexual type of reproduction, you could end up with only great white shark, apex predator type creatures because they're not really required to interact with other entities in a non-hostile manner. You might have to force non-asexual reproduction to achieve higher levels of advancement in the realm of communication, etc.
