[alicebot-aiethics] Alphaville

[smoking pixel]

This doesnt really answer your question directly but it came to mind after 
reading your post.
you wrote "Assume all of that is true [...] I am thinking that the 
omniscience exhibited by the computer in
Alphaville could have a very dehumanizing quality" I agree. For ages god's 
omniscience has been used in an atempt to regulate and control individuals. 
Having such powers _transfered_ to a human made piece of thechno something 
is tremendously more frightening.
To me, being reminded that someone or something can reach you everywhere at 
anytime isnt appealing to me. I'm the kind of person that doesnt like to be 
reminded of how many cameras and records of all sorts are already involved 
in a trip to buy milk and oreos a few blocks away. I treasure my privacy. So 
I think it would be great if everyone could have their personal assistant 
and/or translator of sorts, a link to the eletronic world. Another program, 
maybe a robot or a wristwatch, that would do the job of talking to _the 
voice_ or the ATMs and even toasters, tvs and everything else if we felt 
like it. So instead of a _universal voice_ everyone could have a personal 
_voice_ more or less custom made to our personal needs and likings that 
would be between us and the eletronics. So when we wanted some privacy all 
we would have to do was ask our personal assistant to pull down the shades 
in the bedroom and have it wait in the other room.


----Original Message Follows----
From: "Noel Bush" <noel@alicebot.org>
Reply-To: alicebot-aiethics@list.alicebot.org
To: <alicebot-aiethics@list.alicebot.org>
Subject: [alicebot-aiethics] Alphaville
Date: Tue, 20 Nov 2001 13:27:13 +0300
MIME-Version: 1.0
List-Id: AI Ethics Mailing List <alicebot-aiethics.list.alicebot.org>
List-Unsubscribe: 
<http://list.alicebot.org/mailman/listinfo.cgi/alicebot-aiethics>,<mailto:alicebot-aiethics-request@list.alicebot.org?subject=unsubscribe>
List-Archive: <http://list.alicebot.org/pipermail/alicebot-aiethics/>

Who has seen "Alphaville"?  I just watched this again recently.  In my
opinion this film really has as little to do with artificial
intelligence as it does with space travel or detective thrillers -- but
since these are the surface elements of the movie, if you watch it
expecting something great from that angle you'll be disappointed, since
the story is pretty banal.

However, one thing that does strike me about this film is that it gives
the strong sense that whatever action is taking place is really not
happening in the way you see it.  There are clues to this, such as the
way the characters talk about their use of cars as though it's
interstellar travel.  But beyond all the verbal clues, there's the
SOUND.  The sound (like in a lot of Godard's films) is a very important
aspect.  I am thinking about the constantly recurring "detective"
exclamation figure that pops up in the most incongruent places, or the
way the sound suddenly disappears at some moments, such as when Lemmy
Caution is walking down the hall with Natasha von Braun.  But most of
all I am thinking about THAT VOICE.

THAT VOICE, if I am correct, is Godard's voice, and it's hilarious for
its corrosive, almost lecherous qualities, given that it's supposed to
be the voice of the computer that controls Alphaville.  But more than
these qualities, it's the acoustics of it that really get me.  There's
no "special effect" applied to the voice; in fact, it sounds as if
someone has simply picked up a microphone and is speaking on top of the
movie -- which you can take as being some kind of flaw, or you can
experience as utterly weird in that the voice is so not a part of the
rest of the film's "environment".  The characters in the film hear the
voice and talk with it, but acoustically it is as "present" for us as
for them.  It is somehow emanating from everywhere, nowhere.

Keying in on this about the film, I thought of an ethical question for
us.

We have all had the experience of receiving email from people who are
somewhere else in the world, but still using the same email account.  We
are generally used to this strange quality of Internet communication,
although sometimes it still strikes as odd that the "@" "at" doesn't
really mean "at some location".  Sure, you can say it's "at" some
"location" in a network, but I would bet that deep down none of us still
buy into that "cyberspace" notion whole hog.  It's romantic, okay.

An artificial intelligence, though, could have even less of a sense of,
or care for, the "at".  We will introduce a simple user authentication
scheme for Program D shortly.  This means that no matter where you are,
you can "authenticate yourself" to an Alice bot and pick up a
conversation right where you left it off.  Alice will know that you're a
"Category C" client, that you like pizza, etc.

Now, we can presume that the sorts of face recognition technology
currently being more and more frequently deployed in public places to
thwart terrorism and exhibitionism will at some point be accessible to
mere mortals like ourselves.  So we can guess that at some point in the
near future, you may have to become accustomed to being addressed
personally by vending machines, ATMs, toasters, toilets, and so on.

We can even guess that one or another AI technology may take hold and
become the technology of choice for our upcoming police state.  And our
benevolent administrators may even see fit to use a common personality,
a common voice, a common knowledge set across all manifestations of this
presence.

So we may well be nearing the day when some same voice (I hope it is as
humorous as Godard's) will be talking to us everywhere we go.

The ethical question is this: Assume all of that is true, that it will
happen.  What then are the guidelines for developing our technology so
that it has a good chance of being the technology of choice, but still
retains the characteristics we think are important to avoid
"dehumanization"?  (What are those characteristics?)

I am thinking that the omniscience exhibited by the computer in
Alphaville could have a very dehumanizing quality in its erasure of
place, even moreso than the displacement we already experience by virtue
of our conversion to "Internet geography".

Personally I don't want a large corporation developing the AI of choice,
but I also don't want to see Alice retro-fitted with all of the
"meaning"-bound contraptions that are somehow so appealing to AI
promoters, just in order to win favor.  I think Doug Lenat's assertions
about Cyc's assertions and the "knowledge" they represent are absurd,
and scary.  It makes me think of someone praying to a bronze statue.
("My bronze statue knows the difference between good and evil.")  Alice
suggests a flat, uniform approach to "knowledge" that shuns
metalanguage.  It should keep us on our toes all the time, as much as we
are on our toes with bank tellers and ATMs alike.

The "salvation" of Godard's computer at the end of the film is, of
course, silly.  My wife pointed out that she has seen lots of Soviet
movies that end the same way -- a Soviet hero comes to an imperialist
state where everyone is controlled by a computer, starts reading poetry
or doing something else irrational, and the thing explodes with the
illogic of it all and the day is saved for humanity.

No, even in the sort of future Godard is playing around with, an AI
program is likely to be taken down by some idiotic worm or cracker more
than by some "irrationality bomb".

So I am not asking, "how should a nice computer behave?"  I am asking
about the necessary fundamental characteristics of such a system, in
order for it to be what we want it to be.  Or maybe you disagree, and
you think that an omniscient vocal presence would be comforting, or just
part of the wallpaper.  Hm?

_______________________________________________
alicebot-aiethics mailing list
alicebot-aiethics@list.alicebot.org
http://list.alicebot.org/mailman/listinfo.cgi/alicebot-aiethics


_________________________________________________________________
Get your FREE download of MSN Explorer at http://explorer.msn.com/intl.asp