LISP AI folklore

[Mentifex]

Those rent-a-coder programming wags over at

http://dis.4chan.org/read/prog/1318706528

have launched "Occupy Comp.Lang.Lisp"
with a thread unrighteously entitled,
"Ask John McCarthy anything".

Sorry, script kiddies, v.t.y. Mentifex has already
BTDT (been there, done that) with the following
exchange of views between Mentifex and His Nibs:

http://groups.google.com/group/comp.lang.lisp/msg/04814d7fcec9d16e
asking Dr. McCarthy if he regrets pioneering AI.

http://groups.google.com/group/comp.lang.lisp/msg/5a20daf7c361275b
where the Piafesque answer is, "Je ne regrette rien."

Mentifex (Arthur)
--
http://aimind-i.com
http://mind.sourceforge.net/lisp.html
http://www.tfeb.org/lisp/mad-people.html
http://www.scn.org/~mentifex/AiMind.html

[fortunatus]

Oh, my golly, it's been a long time hearing from Mentifex! How are
you, ol' fella?

[Antti J Ylikoski]

It reads in the Sourceforge site:

"This weblog invites Lisp programmers to implement the main Alife Mind
loop of the simplest artificial intelligence."

I might want to carry out some work for the Mentifex project, if that's
what is desired above. I have been more or less working with LISP since
the 1980's.

The manuscript of my PhD thesis (with some dozen known errors) is in
http://www.tkk.fi/~ajy/diss.pdf . Oh yes, and the work needs to be
carefully scrutinized for correct English.

(BTW in http://www.tkk.fi/~ajy/FreeBASIC/ST.BAS there is there the
famous Star Trek computer game in a modern version.)


kind regards, Mr Antti J Ylikoski
Helsinki, Finland, the EU

[Antti J Ylikoski]

25.10.2011 9:48, Tim Bradshaw kirjoitti:
> On 2011-10-25 02:56:22 +0100, Antti J Ylikoski said:
>
>> Quite on the contrary, I believe that the complexity to create a
>> human-level AI will be enormous, gigantic. The CYC system (Lenat et
>> al.) is a step in the right direction.
>
> Well, as the person you're responding to pointed out, there's this
> awkward thing that that doesn't seem to be the case. There's only so
> much information that is used to get a brain to the point where it can
> start to bootstrap itself.
>

Has anyone attempted to make a cognitive system bootstrap itself by
giving the system in question (very, very) large text corpora -- I
wonder. Of course their contents should be suitable for this purpose.


yours, A. J. Y.

[Andrew Reilly]

Why do you suppose that digestion of text corporea (even complete, if
that were possible) would produce an inteligence that we would recognise
as such? What about real-time stimuli, needs and urges? Body functions?

It's been said that one can't understand a motorcycle from the
description in the manual, unless you have seen one in real life, first.

Cheers,

--
Andrew

[Kaz Kylheku]

On 2011-10-25, Antti J Ylikoski <antti.y...@aalto.fi> wrote:
> Has anyone attempted to make a cognitive system bootstrap itself by
> giving the system in question (very, very) large text corpora -- I
> wonder.

Some academics try this approach on their students, with varying success.

[Antti J Ylikoski]

That is a very good point to my opinion. So we shall need situatedness,
you think, don't you?

andy

[Antti J Ylikoski]

28.10.2011 15:54, Thomas 'PointedEars' Lahn kirjoitti:
> Antti J Ylikoski replied off-topic across 4 newsgroups without Followup-To
> to a posting of Arthur T. "Mentifex" Murray despite being notified several
> times before:
>
>> […]
>
> Which part of "Do not feed the troll" do you need tatooed on the container
> of your malfunctioning brain simulation?
>
> *plonk*
>
>
> F'up2 set appropriately

See my entry about "Kritik der reinen Mentifex".

I wanted to see this program and its world. It is not real good novel
research, after getting to know what it contains... But I just wanted to
see this thing, is it good for anything. As I said, like a new Prolog
interpreter.

Building a model of the human brain after Aristotle is not a good
scientific idea. It's like making skyscrapers after Egyptian pyramids
(there is one I recall in Los Angeles.....)

andy