ianparker2@[EMAIL PROTECTED]
wrote in news:1117108869.710819.123360
@[EMAIL PROTECTED]
> The main argument against Evolution is the failure to produce AI.
> Intelligence has proved too complex to be generated in a random way and
> too complex even for us to understand.
>
> Evolution isn't just a biological concept, it, in the shape of Genetic
> Algotithms, is a mathematical and computational concept. Ee can now ask
> the question as to whether Evolution could have occured randomly. My
> own investigations cast doubt on whether scientists really believe in
> random evolution.
>
> An army of PhDs is working on AI. They are all working on specific
> projects in weak AI. No respectable peer group believes either in
> strong AI, or that the route to objectives in weak AI is through
> Genetic Algorithms.
This is of course an argument in favor of evolution - intelligence can
obviously not be produced by design, since all these PhD's can not
produce even weak AI by design.
> If Intelligence is the result of random mutations it follows that
> strong AI could be produced on a grid or a BOINC by an analogous
> process. Who believes thiis in the sense of being prepared to put their
> money where their mouth is.
Since current computational devices are several orders of magnitude less
complex than human brains, I fail to see why one should expect them to
exhibit intelligence on a human level. I would expect them to be able to
exhibit intelligence on the level of an ant (with enough millions of
iterations) - and I think that current algorithms are roughly in this
category, albeit with little of the robustness exhibited by evolved vs.
designed systems.
> I also feel that by not proposing alternatives to Darwinism a dangerous
> myth is created. The myth that Evolution is a mathematically efficient
> process. Evolutionalists are saying that Evolution is better than what
> the army of PhDs is doing. I feel that the priority should be the
> teaching of good Numerical Analysis, that is if your priority is
> keeping up with China and India in the scientific stakes.
I am unaware that anyone has argued that evolution is a mathematically
efficient process. And if you want to teach good Numerical Analysis you
would be hard pressed to find a better example than Fisher.
Yours,
Bill Morse
|
