Re: Human mind Turing computable according to Eliezer

From: Wei Dai (
Date: Fri Oct 08 2004 - 06:10:25 MDT

On Thu, Oct 07, 2004 at 12:39:19AM -0400, Eliezer Yudkowsky wrote:
> Mitchell Porter wrote:
> >
> > Look, do you agree that blue patches, loud noises, etc., exist, or not?
> > If yes, explain what's blue about a bunch of cortical neurons. If not,
> > explain how I can interpret my experience without blueness, loudness,
> > etc. In my world, these things are *epistemically* basic, whether or
> > not they're ontologically basic.
> I'm replying to Mitchell about this off-list; we'll see how that goes.

Why the off-list reply?

Let me try this approach to the issue, and see if it's helpful. Consider
the following thought experiment. You're a disembodied brain living in a
jar, with remote control of a telepresence robot. You just heard that the
storage warehouse housing your jar is on fire, and rush the robot back
to save yourself. The bad news is there are thousands of brains in the
warehouse, and you don't know which one is yours. The good news is there's
a brain scanning machine with resolution down to individual neurons, and
you can try to use it to find out which brain is you. Question is, how to
do it?

How do you take a bunch of neurons and figure out what it's experiencing?
Would the answer to this question also answer Mitchell's question, or is
he asking something else?

Is this question worth trying to answer? A mind with full access to its
self-state presumably wouldn't have this problem with determining which
brain is itself. It would just directly compare the scanner output with
its own self-inspection. Assuming that we'll get this feature soon,
is there some other scenario where the answer would be useful?

This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:49 MDT