|Original post by ragamuffin|
|Original post by The Frugal Gourmet|
For instance, although genetic algorthims or neutral networks computers may *rely* upon mathematical computation at some level, they do not necessarily generate their solutions algorithmically. They can actually generate solutions to problems *associatatively*.
I do not think Goering's proof positively shows that solutions to the aforementioned problems cannot be generated through an associative process.
i dont understand what you mean? genetic algorithms are algorithms, although they do not generate their solutions symbolically, they still use an algorithmic process to evolve an answer. or am i missing something here?
I absolutely agree. When you get right down to it, a genetic algorithm still relies upon an algorithm or computational/mechanical process.. but I'm not convinced it matters.
It seems to me that Goering's proof only shows us that an algorithm cannot be the direct process used for determing a solution. I am not still convinced that you cannot use algorithms/mechanics/math to build a system that itself produces answers *non-algorithmatically*. Whatever the system is built upon seems irrelevant to me. As long as you the system spits out answers using a process that isn't ITSELF an algorithm, then it can indirectly be built upon all the algorithms it wants and still not fit this description.
In short, Goering's proof shows us that math alone cannot solve all problems the human brain can solve. It does *NOT* however show us that math/mechanics cannot build something that *can*.
I still believe human beings are simply mechanical, when you get right down to it, and are composed of the same set of chemicals, respond to the same physics and math as everything else in the universe.
You said "although they [genetic algorithms] do not generate their solutions symbolically, they still use an algorithmic process to evolve an answer.", and I don't see why the latter matters. I don't think Goering shows us 'algorithms are never used by intelligent beings'... I think Goering just shows us 'algorthms are not exlusively used by intelligent beings to solve problems.
However, I will admit to not being an AI expert, and the reason I am chiming in here is solely for my own intellectual curiosity. I am very interested in learning why AI is not possible with computers... but I am not yet convinced by this Goering proof. I don't think it satisfies all the questions.[Edited by - The Frugal Gourmet on August 25, 2004 3:54:03 PM]