Cultimo on Tue, 26 May 1998 18:21:09 +0200 (MET DST)


[Date Prev] [Date Next] [Thread Prev] [Thread Next] [Date Index] [Thread Index]

Re: <nettime> Sentient Experiment #1.1


>Sentient experiment log #1.1
>
>With drunk: Test in the destruction of the direction by means of repeated
>machine translation.
>
>Method: The expression on several occasions was translated English-French
> and French-English babelfish of the use.
>
><http://babelfish.altavista.digital.com/cgi-bin/translate>
>
>Result:
>
>1. we no longer have routines we have acronyms
>
>2. nous n'avons plus des sous-programmes que nous avons des acronymes
>
>3.we do not have any more of the subroutines which we have of the 
>   acronyms
>
>4. nous n'avons pas davantage des sous-programmes que nous avons des
>    acronymes 
>
>5. we do not have more subroutines which we have of the acronyms 
>
>6. nous n'avons pas plus de sous-programmes que nous avons des
>    acronymes
>
>7. we do not have more subroutines than we have acronyms 
>
>Conclusion: machine translation a suitable process for the destruction of the
>direction
>
>Let us consider pour Davantage de Recherche: Troislangagesdans
>Laumdrehung Essayez de Traduire average.

So in short: one can conclude that a machine has a big problem with the 
grammar of denial. And perhaps one can conclude that a machine has a 
problem with the philosophical implication of what is not. Next question: 
What does a machine do with possibilities?

Someone once tried to convince me that we - humans - cannot count. We can 
tell stories, thats all we can. And a machine - a machine can count. 
Based on that it even can play checkers (even though some people stated a 
great many years ago that would be impossible). Telling stories and using 
language is about imagening that which is not. Is translation the final 
border: will it ever be possible to teach a machine to translate? To tell 
stories? In an essay 'The ghost and the machine' written by Italo Calvino 
he stated that both the structuralistic aproach and the Freudian approach 
of stories and dreams add up to the same: the combination of a restricted 
group elements into endless combinations. (structuralistic: a story is 
formed on grammar bassed on positions of actors, freudian: dreams (and 
stories) are formed from symbols comming from our subconscious)

Hence: it will be possible to teach a machine how to tell a story. It 
only has to learn how to combine the ellements that can possibly form a 
story. 

At the M.I.T. they once made a program called 'the endless conversation': 
based on a database a  machine made a conversation from the elements 
there in. All statements he could use where labelled: question, awnser, 
reaction. And they where labelled by subject. So if the machine choose a 
random question about choclate milk it searched for an awnser and a 
reaction about that subject. The conversation went on until someone 
stopped the program. 

So we can teach a machine how to combine elements. But we will never 
teach a machine the function and meaning of fantasy. Because fantasy has 
to do whith that which is not and a machine can only work with that which 
is countable.

Jeroen Goulooze ;-)
---
#  distributed via nettime-l : no commercial use without permission
#  <nettime> is a closed moderated mailinglist for net criticism,
#  collaborative text filtering and cultural politics of the nets
#  more info: majordomo@desk.nl and "info nettime-l" in the msg body
#  URL: http://www.desk.nl/~nettime/  contact: nettime-owner@desk.nl