From: Jeff L Jones (email@example.com)
Date: Mon Mar 24 2008 - 16:14:41 MDT
On Mon, Mar 24, 2008 at 1:00 PM, J. Andrew Rogers
> On Mar 24, 2008, at 12:06 PM, Jeff L Jones wrote:
> > I still don't understand why you guys are talking about GLUTs.
> > They're completely and entirely impossible, even in prinicple.
> The same is true of Turing Machines.
Good point, although I feel that Turing Machines are a lot more useful
to think about than GLUTs. They can answer all sorts of questions
about what problems can or cannot be solved... since they are the
infinite limit of a series of finite machines that *can* be built, and
are physical. GLUTs on the other hand aren't the infinite limit of
anything that could be used to pass a Turing test, or could be said to
"implement consciousness". I guess you could argue that finite tape
Turing machines, if they are thought to be literally a piece of paper
sliding through a device, will always be way too slow to pass a Turing
test. But I don't think the piece of paper sliding through a device
is the essence of what the Turing machine is about... I don't think
it's unreasonable to believe you could make one fast enough to
implement consciousness if you used nanobots for the device and giant
strings of molecules for the tape. The tape size would be finite, but
other than that it would really *be* a conscious Turing machine. I
don't see any analogue of that for GLUT's.
> I have always liked the GLUT as a device because it forces people to
> acknowledge, against their intuition, that there is no magic hiding in
> the machinery.
I admit, that it does have this appeal.
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:01:02 MDT