Re: Another Introduction

En réponse à Alex Fink :


> >But have you thought about a compression method that works on the bits
> >alone, before the mapping to syllables? :-)
>
>It seems like this would make it even more impossible to use!  No, I
>haven't, but I'll have to look into it... :-)
It occured to me that this could be a form of Speedtalk for intelligent
machines. A robot would create the utterance in its computerised "mind" in
the binary language (the normal way a computer handles data), compress it,
map the result into phonemes and speak it. The receiver would listen to the
utterance, convert it back into binary form, unpack it and parse and
understand it. Since most would happen in their "minds", i.e. a
supercomputer, it would be quasi-instataneous for human beings and thus
robotic conversations would just sound like human conversations, except
that they would be usually shorter, and unintelligible without a
translation machine.

Of course, such machines would probably rather connect each other directly
(by radio for instance, or through a physical line) rather than resort to
such a slow interface as talking. But if the conditions made those
connections impossible (jamming of radio frequencies or practical
impossibility to physically connect), it would be a backup communication
medium (or the programmers could have programmed the machines to use speech
as their primary mode of communication. Nobody's perfect ;))) ).

Christophe Grandsire.

http://rainbow.conlang.free.fr

You need a straight mind to invent a twisted conlang.