tedd <····@spamcop.net> wrote on Tue, 13 Aug 2002:
>Albert van der Horst wrote:
>>
>> To qualify as an AI programming language, a system written in
>> that language must be able to program itself, i.e. write its
>> own programs, then execute them.
>> As such LISP, FORTH and (far out) assembly, shell scripting qualify.
>> None of the compiled languages (FORTRAN PASCAL C C++ Java) fit
>> the bill.
>
> Albert:
>
> I wasn't aware that an AI programming language was defined by
> its ability to write its own programs and then execute them.
And nobody else is aware of such a definition, because it's not
what Albert van der Horst wrote. He wrote that the _system_ (e.g.,
http://www.scn.org/~mentifex/mind4th.html -- Mind.Forth Robot AI)
programmed in the language "must be able to program itself."
Mind.Forth itself, the humble but by default most advanced AI,
has only only two weeks ago learned to learn by asking questions,
and has only two days ago learned to answer "why" with "because...."
But eventually Mind.Forth or its derivatives will program AI Minds.
> So, you mean that rule based systems, inference engines,
> genetic algorithms and the list goes on and on -- are not AI?
Strictly speaking, such things are not artificial intelligence (AI)
if they are not able to _think_ and ask questions and learn, as can
http://www.scn.org/~mentifex/jsaimind.html -- the Tutorial AI Mind.
> Because, I can write those types of programs in almost any language.
>
> tedd
You forgot about mutation and natural selection.
I think those are the keys to development of intelligent entities.
·····@victoria.tc.ca (Arthur T. Murray) wrote in message news:<········@news.victoria.tc.ca>...
> tedd <····@spamcop.net> wrote on Tue, 13 Aug 2002:
> >Albert van der Horst wrote:
> >>
> >> To qualify as an AI programming language, a system written in
> >> that language must be able to program itself, i.e. write its
> >> own programs, then execute them.
> >> As such LISP, FORTH and (far out) assembly, shell scripting qualify.
> >> None of the compiled languages (FORTRAN PASCAL C C++ Java) fit
> >> the bill.
> >
> > Albert:
> >
> > I wasn't aware that an AI programming language was defined by
> > its ability to write its own programs and then execute them.
>
> And nobody else is aware of such a definition, because it's not
> what Albert van der Horst wrote. He wrote that the _system_ (e.g.,
> http://www.scn.org/~mentifex/mind4th.html -- Mind.Forth Robot AI)
> programmed in the language "must be able to program itself."
>
> Mind.Forth itself, the humble but by default most advanced AI,
> has only only two weeks ago learned to learn by asking questions,
> and has only two days ago learned to answer "why" with "because...."
> But eventually Mind.Forth or its derivatives will program AI Minds.
> > So, you mean that rule based systems, inference engines,
> > genetic algorithms and the list goes on and on -- are not AI?
>
> Strictly speaking, such things are not artificial intelligence (AI)
> if they are not able to _think_ and ask questions and learn, as can
> http://www.scn.org/~mentifex/jsaimind.html -- the Tutorial AI Mind.
>
> > Because, I can write those types of programs in almost any language.
> >
> > tedd