[ExI] Is the brain a digital computer?

[Stefano Vaj]

On 24 February 2010 13:24, Stathis Papaioannou <stathisp at gmail.com> wrote:
> The physical Church-Turing thesis has not been proved. Non-computable
> functions exist as mathematical objects, but it is not known if they
> play a role in physics, or in the physics of the brain in particular.

I take "non-computable" in this sense simply to mean that there are no
algoritmic shortcuts, and that you have to run the system (or any
emulation thereof) to see where it leads.

Remaining in the field of cellular automata, a few instances of such
scenario are easily found.

But this does mean that the final outcome of step "n" cannot be
determined by any universal computer which goes through the very same
steps. Very possibly, with a definite loss of performance.

In this sense, I think one could be reasonably argue that a human
brain might be the most efficient way to produce a human identity.
Many AGI partisans take on the contrary a little too much for granted
that an electronic computer could easily compete with it...

-- 
Stefano Vaj