Re: burte force and knowledge

[David Mechner <mechner@xxxxxxxxxxx>]

> Has this thread degenerated to dictionary flames?
I didn't mean it as a flame, just an explanation of the (apparent)
disagreement. 

> I believe all 361 are evaluated simultaneously and sunconsciously 
> and the best ones pop into consciousness.

I disagree, but it doesn't really matter whether they're evaluated
totally in parallel or just quickly in serial - in either case we agree
an important part of the initial evaluation is "subconscious" - not
accessible to introspection. 

I also agree with your conclusion:

> If go had some trivial solution then it would surely have been
> found by now. Go is hard, and that implies taht a substantial 
> amount of processing is going on internally when humans play.  
> We shouldn't expect computers to do well until they do a 
> similar amount of processing.

Excellent point. However, what does similar mean? It's awfully hard to
say how much processing we're doing when we play go, and also
practically impossible to meaningfully compare "amount of processing"
between brains and conventional computers. And then, even if we could
somehow, the architecture of the brain biases the kinds of solutions we
find to a problem. There may be more efficient ways of solving the
problem given the capabilities of serial computers.

In any case, the main concern isn't (shouldn't be) getting the right
amount of computing power into the problem any old way, but using the
computing power we have as efficiently and effectively as possible.

-David

-- 
David A. Mechner            Center for Neural Science
mechner@xxxxxxxxxxxxxxxxx         4 Washington Place, New York, NY 10003
212.998.3580                http://cns.nyu.edu/~mechner/