Some Ideas to share

50 views
Skip to first unread message

ssj5q...@gmail.com

unread,
Jul 16, 2020, 1:39:01 PM7/16/20
to opencog
My username in the chat was "foobaring foobaring" and still is, of the opencog conf

I would like to share with you my research on the subject. I saw Dr. Ben podcast with lex fridman.
And I think that your ideas are close to what I was doing.

So to explain better what I was talking about, I prepared this image.

I would like to know what your thoughts are. And if it is helpful.
AGIAlgorithmv1.png

Linas Vepstas

unread,
Jul 16, 2020, 2:32:34 PM7/16/20
to opencog
Hi foobar!

You might enjoy reading about "model theory" which deals with much of what you talk about. Sorry, I do not have any good references I can point you at.  (the wikipedia article on it is terrible).  As a general note, the newer the text the better (this stuff is changing rapidly)  You might also like reading about "relational algebra". (the newer the better; the old stuff has poisonously-ugly notation that will make you think the wrong ways about the wrong things for the wrong reasons)

--linas



--
You received this message because you are subscribed to the Google Groups "opencog" group.
To unsubscribe from this group and stop receiving emails from it, send an email to opencog+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/opencog/07cb4a18-ce90-4920-b0f4-409bf16f172fn%40googlegroups.com.


--
Verbogeny is one of the pleasurettes of a creatific thinkerizer.
        --Peter da Silva

Linas Vepstas

unread,
Jul 16, 2020, 2:34:39 PM7/16/20
to opencog
Oh, I forgot to add: whatever you find, its important to read as much of that as you can tolerate 1 or 2 pages or 10 or 20  or all of it. If its too hard or too boring, just move on to something else. Go towards it from many different angles.

--linas

ssj5q...@gmail.com

unread,
Jul 16, 2020, 3:20:30 PM7/16/20
to opencog
Thanks. I will try to keep up with the theory.

ssj5q...@gmail.com

unread,
Jul 16, 2020, 7:11:55 PM7/16/20
to opencog
I'll try to explain myseft some more. And that will be all my attempt to help.

What I know about your work with opencog is that you are trying to do an AGI with, amount others things, a knowledge graph 
and a set of algorithms. Is that correct? It may be an over simplistic version of what your doing. But you can correct me if i'm wrong.

I was doing that too. And what I have share with you is a sort of example of what I am doing.
I am also using a knowledge graph and algorithms. The graph if is very simplistic, I use 'equal' (=) to connect nodes and groups to 
represent complex data. And then the algorithm is just permutation basically. And the data have not type, they are just strings.

I hope someone get something out of it. And thanks for your time.
Reply all
Reply to author
Forward
0 new messages