[Wittrs] Re: On the Mechanism of Understanding
- From: kirby urner <kirby.urner@xxxxxxxxx>
- To: Wittrs@xxxxxxxxxxxxxxx
- Date: Thu, 13 Aug 2009 22:58:11 -0700
On Thu, Aug 13, 2009 at 6:26 PM, iro3isdx<xznwrjnk-evca@xxxxxxxxx> wrote:
>
>
> --- In Wittrs@xxxxxxxxxxxxxxx, kirby urner <kirby.urner@...> wrote:
>
>> I thought the Emperors New Mind (Sir Roger) was pointing in a good
>> direction in some ways.
>
> It's mostly just silly (in my opinion).
>
>> I like the idea of "non-computable".
>
I'll cop to not reading Emperors New Mind cover to cover.
We've had the good fortune in Portland, Oregon to have hosted Sir
Roger a number of times, so I've heard him lecture in person on these
and other themes, sometimes with a more intimate dinner after, more
questions after drinking (so better answers sometimes, at least from
the listener's point of view).
He also came to our Math Summit in 1997, organized by this think tank
of ours (ISEPP).
> Perhaps it has a curious ring to it. However, anything non-computable
> is infinite (whether a number with infinite decimal expansion, or a
> procedure that doesn't finish after a finite number of steps).
>
The way Penrose pictures it is more like making a conceptual leap,
connecting two or more dots in some inscrutable way that we could
never hope to "program" (as in "write an algorithm to do"), but then
sometimes, once the dots are connected, we go back and prove the
connections by rational steps.
It's like the mind "flies to the destination" (by a non-computable
leap) and then we put in a paved road by means of reason.
It's really easy to stump Big Blue with made up chess puzzles that
even a child could solve.
We'll never be able to program genius.
Artificial intelligence will always be just that: artificial.
That would seem to be the gist of his view, though I'm open to other
interpretations (including from Sir Roger of course!).
> We don't live long enough for infinite things to crop up. And that's
> why Penrose's ideas are just silly.
>
> Regards,
> Neil
>
I didn't get the impression "infinity" was that important in his
thinking around this topic, but as I've said, I've not read the book
cover to cover.
I think a Finite Universe model (per Knuth's for example) could just
as well host non-computable events.
Kirby
Other related posts:
