Extrapolating First Working AGI MindForth

53 views
Skip to first unread message

menti...@gmail.com

unread,
Jul 9, 2018, 5:16:30 AM7/9/18
to
Artificial General Intelligence (AGI) has arrived in MindForth and its JavaScript and Perl symbionts. Each Mind is expanding slowly from its core AGI functionality. The http://ai.neocities.org/MindBoot.html bootstrap sequence of innate concepts and ideas can be extended by the machine learning of new words or by the inclusion of more vocabulary in the MindBoot itself.

We may extrapolate from the current MindBoot by imagining a Forth AI that knows innately the entire Oxford English Dictionary (OED) and all of WordNet and all of Wikipedia. Such an AI would be well on its way to superintelligence.

If there is no upper bound on what a Forth AI may know innately, why not make full use of Unicode and embed innately the vocabulary of all living human languages? Then go a step further and incorporate (incerebrate?) all the extinct languages of humanity, from LinearB to ancient Egyptian to a resurrected Proto-European. Add in Esperanto and Klingon and Lojban.

Mentifex
--
http://agi.topicbox.com/groups/agi
http://ai.neocities.org/mindforth.txt
http://cyborg.blogspot.com/2018/07/clpm0709.html
http://groups.google.com/d/msg/comp.lang.forth/ZLpZEnA-iGk/y6Nx1DV8BwAJ

NN

unread,
Jul 9, 2018, 9:59:38 AM7/9/18
to
I was looking at : https://ai.neocities.org/mindforth.txt and I wondered if you have ever considered ;

1) separating the comments out or deleting the unnecessary ones ,
2) refactoring so that words are smaller eg InStantiate , mindboot etc

It's a fairly large program and not always easy to step through.

menti...@gmail.com

unread,
Jul 9, 2018, 4:11:02 PM7/9/18
to
On Monday, July 9, 2018 at 6:59:38 AM UTC-7, NN wrote:
> On Monday, 9 July 2018 10:16:30 UTC+1, menti...@gmail.com wrote:
> > Artificial General Intelligence (AGI) [...]
>
> I was looking at : https://ai.neocities.org/mindforth.txt and I wondered if you have ever considered ;
>
> 1) separating the comments out or deleting the unnecessary ones ,

As the author of MindForth, I am constantly deleting commented-out diagnostic messages after they have served their purpose. I do leave them in through one or more releases so that programmers may see what is happening in the code.

Anyone is welcome to take the pre-existing code and start a fork as a branch of AI evolution. Anyone studying the code may compare it with the JavaScript and Perl versions, which are the same program.

> 2) refactoring so that words are smaller eg InStantiate , mindboot etc

http://ai.neocities.org/InFerence.html -- has a list of the "words" or modules at the bottom of the page. There are about forty of them and most of them have been factored down to bare essentials.

http://ai.neocities.org/InStantiate.html -- has a lot of confusing things going on, but all dedicateed to the task of creating a new associative "node" or "instance" of a concept at a given time-point. I suppose InStantiate could be factored down into one Forthword for internal thinking and another Forthword for external input.

If anybody still runs Forth to control robots, it would be nice to see someone put some robotic control code into the http://ai.neocities.org/Motorium.html module with a
repetition-threshold trigger in the http://ai.neocities.org/Volition.html module -- so that any Forth robot that thinks to itself repeatedly "I want to rotate (my body)" will actually start to rotate by a robotic act of free will.

It would be nice to see MindForth run in various versions of Forth and on platforms like the GreenArrays chips.

>
> It's a fairly large program and not always easy to step through.

Thanks for looking at MindForth.

Cheers,

Arthur
Reply all
Reply to author
Forward
0 new messages