From: Jochen.Fromm@t-online.de
Maybe the secret of AI is not to use AI in games,
but to create true AI with the help of games.
What I mean is not artificial intelligence
on the level of "The Sims".
In "The Sims", each world object is a stand-alone
code object or class which contains a list of desires
it can satisfy. If a Sim character walks near
an object, let us say a frigde, he hears the
fridge say it "can satisfy hunger". The object
also contains the data on how to use it - how
to open and close it, how to switch it on, etc.
Although a "Sim" looks like it understands the
world, it is a stupid as it can be. It relys on
the objects in the world to tell it what it can
do and how it should do it.
The question is : is it possible to create an
intelligent character, who really understands
a simulated 3D world - a world which looks like
the real world ?
I think it is possible. H.A. Simon says in
"The Sciences of the Artificial" :
"A man, viewed as a behaving system, is quite
simple. The apparent complexity of his behavior
over time is largely a reflection of the complexity
of the environment in which he finds himself."
The structure of an old tree is very complex,
but it can by described by a simple fractal or
a simple iterated function system - by rules
to let it grow. I guess the secret of artificial
intelligence lies here :
* create a adaptable or adaptive system
(with the right kind of internal architecture)
* place the system in an artificial environment
which is complex enough
* let the system learn and grow
At least this is what humans do for nearly 20
years : to grow and to learn. Of course
the world must be complex enough to create
a complex character. For example an online
multi-player game in a true 3D environment.
Today the characters in these games are still
rigid and fixed, with only one or two face
expressions and rigid movements.
But progress doesn't stop, and in a few years
these systems will look like the real world
What today is possible in films like the Matrix
will be available in online computer games in a
few years. The tricks and effects from early
films - for example "Tron" - can be easily
reproduced today.
A large percent of the world we recognize is
made up of other persons and interactions with
them. We recognize other people through their
look and especially their face. Face recognition
alone is a difficult task for its own. A single
computer will be overcharged with the task
of simulation of all other players and their
representation and regcognition in one special
player. But with the help of online 3D games
you are possibly able to circumvent this
problem.
The good thing is, if you have created a system
which understands a 3D world, you have created a
system which can understand natural language, since
language is a description of the physical, real
world and is inseparable from it. As soon as
you start to separate words from it though,
they loose their meaning, and you can regain
it only if you pull them "back to earth" with
metaphors. As George Lakoff said :
"As soon as one gets away from concrete physical
experience and starts talking about abstractions
or emotions, metaphorical understanding is the norm."
To create true AI, you have go back to "concrete
physical experience". So either you build robots
(bring the computer in the world), or you create
simulations (bring the world in the computer).
The most complex robots function only on video
tape or if you don't look. It is easier to create
simulations.
Will it be possible to create true AI ? If it
is, will it be the next major step of evolution ?
What do you think ?
[ comp.ai is moderated. To submit, just post and be patient, or if ]
[ that fails mail your article to , and ]
[ ask your news administrator to fix the problems with your system. ]
--- SoupGate-Win32 v1.05
* Origin: you cannot sedate... all the things you hate (1:229/2)
|
