Chopped Liver.
~PM.
From: [email protected]
To: [email protected]
Subject: RE: [agi] Re: Uber big scary monster OpenCog diagram
Date: Tue, 20 Nov 2012 20:17:54 -0800
Ben,
I like your new diagram and it's very clear to see the components now. I think
your architecture is in a good direction.
I think having an elemental representation unit is a good start: i.e., using
ATOMS in OpenCog or MONADS in PAM-P2.I also think having multiple agents /
processes act upon this working memory of elemental units is viable, scalable,
and necessary.There are differences between OpenCog atoms and PAM-P2 monads.
For example, PAM-P2 does not use probablities in themonads, they are purely
temporal and structural. We differ in what we conceive the elemental units to
be, and what processes act upon them.
We both recognize that there needs to be clustering/induction, analogy,
perception, LTM versus STM, planning, etc. But PAM-P2 employs other processes
like correlators, regulators, compensators, where Open Cog has dimensional
embedding spaces, and a world simulation engine.
But to integrate all of these aspects into a large system is the challenge and
not many designers accept. Some people feel that AGI will occur through a
single algorithm or very few algorithms. I disagree, I think it's going to be
an aggolmerationof highly specific processes synthesizing and merging concept
units in a heterarchical repository.
Whether we build a Volkswagen, Rolls Royce, or Suburban, the cognitive system
will have similar features, four wheels some number of doors, and come with or
without air-conditioning. Our similar design constraints and assumptions will
necessarily lend itself to a particular architectural result.
I don't know what this architectural approach or school is called, perhaps this
is still patternist, but I think it could be called something else. I believe
PAM-P2 is constructivist because the monads and schemes in PAM-P2 are being
constructed by the agent processes.
So much for nomenclature or categorizations.
Good diagram.
Cheers!
Michael Miller
(Piaget
Modeler)------------------------------------------------------------------------------------------------------------------------------------------------Confidential
- This message is meant solely for the intended recipient. Please do not copy
or forward this message without the consent of the sender. If you have received
this message in error, please delete the message and notify the sender.
> Date: Wed, 21 Nov 2012 10:07:03 +0800
> Subject: [agi] Re: Uber big scary monster OpenCog diagram
> From: [email protected]
> To: [email protected]
>
> And finally, the SmartDraw version of the big scary OpenCog conceptual
> diagram that I posted (in rough hand-drawn form) a couple months
> ago...
>
> http://goertzel.org/MonsterDiagram.jpg
>
> On Thu, Sep 13, 2012 at 4:26 AM, Ben Goertzel <[email protected]> wrote:
> > For those who like ugly pictures... ;)
> >
> > Here is a diagram, scribbled by me, showing some bits and pieces of
> > what the major parts of OpenCog do, in the specific context of
> > carrying out some simple physical and conversational actions involving
> > building stuff with blocks.
> >
> > It's extremely oversimplified, as it shows a couple dozen Atoms and
> > MindAgent-actions, in a situation where there would be at minimum
> > hundreds of thousands of Atoms and dozens to hundreds of MindAgent
> > actions per second. So it's purely intended to be didactic and
> > evocative.
> >
> > It'll get re-drawn in SmartDraw sometime in the not too distant future...
> >
> > -- Ben
>
>
>
> --
> Ben Goertzel, PhD
> http://goertzel.org
>
> "My humanity is a constant self-overcoming" -- Friedrich Nietzsche
>
>
> -------------------------------------------
> AGI
> Archives: https://www.listbox.com/member/archive/303/=now
> RSS Feed: https://www.listbox.com/member/archive/rss/303/19999924-5cfde295
> Modify Your Subscription: https://www.listbox.com/member/?&;
> Powered by Listbox: http://www.listbox.com
AGI | Archives
| Modify
Your Subscription
-------------------------------------------
AGI
Archives: https://www.listbox.com/member/archive/303/=now
RSS Feed: https://www.listbox.com/member/archive/rss/303/21088071-c97d2393
Modify Your Subscription:
https://www.listbox.com/member/?member_id=21088071&id_secret=21088071-2484a968
Powered by Listbox: http://www.listbox.com
