X-Message-Number: 11516
Date: Mon, 05 Apr 1999 00:09:36 -0700
From: Mike Perry <>
Subject: Simulated (really Emulated) Persons

Note: this message was originally sent several days ago, as

indicated. It doesn't seem to have arrived (as of April 4) so I am resending it.

Date: Thu, 01 Apr 1999 01:37:47 -0700
To: CryoNet
From: Mike Perry <>
Subject: Simulated (really Emulated) Persons

Some thoughts on Thomas Donaldson's posting, #11491:
>
>And on simulation versus reality, for Mike Perry: I will ignore the issue
>of quantum behavior, mainly because everything I know about how our brain
>works suggests that quantum theory plays the same role in our brain as it
>does in a diesel engine: it's usually irrelevant. 
>
It does seem irrelevant, true, but we haven't achieved an emulation of human
or humanlike consciousness in a computer yet either. So it remains to be
seen what sort of device that will take, and maybe the most practical route
will be through quantum devices.

>It is not even clear that the thoughts of a simulated human should count
>as real thoughts. Basically, simulation involves the construction of 
>a computer program (or object) which imitates a real object SYMBOLICALLY.
>No one believes that they will be seriously hurt if they crash a simulated
>airplane; the basic point Searle was making was that we can play with
>symbols according to various rules all we want, but those symbols (and
>our play) says nothing about the world unless we explicitly AND SEPARATELY
>identify the behavior of the symbols with something about the world.

By a reasonable thought-experimental license, we could imagine an *emulated*
human who didn't know he/she was "only an emulaton" and for whom pain, fear,
etc. were just as "real" as for a "real" human. Again, maybe this could only
happen, practically speaking, in a quantum device. Seth Lloyd's universal
quantum simulator, however, would seem to open the theoretical possibility.
(See Lloyd, S. "Universal Quantum Simulators," Science 273 1073-1079 [23 Aug
1996].) To me it seems specious or mystical to argue that a thinking being
emulated at the quantum level would not have "real" thoughts, but whatever
you think on this subject, if such a being produced intellectual property it
would be real. So once again, while a simulated stomach would not do real
digestion, a simulated poet *would* do real poetry.

>(Yes, I've pointed out before that a symbolic system just might be so
>extensive and so complex that it could only be identified with the real
>world in one way --- but proving that claim by experiment takes far more
>than simply asserting it to be true). 

I have some thoughts on this in my book. Informal arguments favoring the
universal language are fairly straightforward; some of them have been
considered here before.

>If we do not carry out that identification, or perhaps carry it out
>wrongly, then we are only playing games with symbols. Any conclusions we
>reach mean nothing. Not only that, but if a simulated human being has
>simulated thoughts, those thoughts themselves mean nothing unless a REAL
>human being (or other REAL CREATURE) identifies them with thoughts and
>assigns meaning to the trains of symbols of which they consist.

No, here is where I disagree. I think processing of "symbols" can embody
intrinsically experiences, meaning, etc. No outside "interpreter" is needed.
Consider this: we have certain "real" processes going on right now, by which
we live, move, and experience all we experience. These processes amount to
processing at the quantum level. If instead we emulated these processes in
Lloyd's universal *quantum* simulator, are we to suppose that they would no
longer have an intrinsic "realness" but must now be "interpreted" by some
outside agent to have meaning? To me, it doesn't seem so. I think creatures
in the simulator would have just as "real" experiences, subjectively, as
their counterparts on the outside. Otherwise it seems we are into mysticism.

However, in spite of all that, what you say is not completely off-base. I
would take a hard look, though, at "those thoughts themselves mean nothing
...". Mean nothing to whom? I would say, to any "REAL CREATURE" unless THAT
CREATURE "identifies them with thoughts and assigns meaning ...". This
identification is easy and automatic if you are one of these "real
creatures" and you are watching and interacting with another one who is like
you. So you don't question that the other one is having "real" experiences
too. But if you try to imagine yourself interacting with a strange process
in a possibly strange way, it doesn't seem like a "real" person and you
question whether it could be. But again, in my view, whether it would be
would depend only on certain intrinsic properties of the process rather than
its embodiment in a particular type of construct or device.

Mike Perry

Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=11516