Mike, I think you may have narcissist tendancies, and may lack the ability to
empathize or even simulate.
As it stands, I think the only thing that may satisfy you is a working system,
and even then you'd probably search for a homunculus within it.
When PAM-P2 is ready, I'd be happy to demonstrate it to you, until then, you
can find some papers on my site.
Best.
~PM.
From: [email protected]
To: [email protected]
Subject: Re: [agi] Re: Superficiality Produces Misunderstanding - Not Good
Enough
Date: Tue, 23 Oct 2012 18:37:42 +0100
PM,
You do understand that that’s a total non-answer? “We will give this
problem to the computer, and hopefully the computer will solve it..”
That’s waffle. The computer has the same problem as you or any agent
-
HOW will it solve the problem? -
what are the COMMON ELEMENTS - and COMMON RELATIONSHIPS OF THOSE
ELEMENTS – that will enable you or the computer to identify these different
figures as belonging to the same class of “chair” and not “collages of
wood” or “piles of assorted forms” or “computer desk” or “collections of
tools”?
ARE there any common elements?
You haven’t identified any.
[I am using both you guys here as representative figures – it isn’t
personal – what you are doing, everyone is doing – totally evading the problem
and any analysis of the problem. And yet waffling on confidently about how of
course semantic nets or some other form of narrow AI will definitely solve the
problem. B & B are saying in effect that those chairs share a “pattern”. But
nothing being mooted works – or is in any way relevant to the
problem.
This IS the problem of AGI – whether you define it as
creativity/concepts/visual object recognition/metaphor – the chair problem I
have set can be reframed to fit them all.
Do you want to face it or evade it?
From: Piaget Modeler
Sent: Tuesday, October 23, 2012 6:14 PM
To: AGI
Subject: RE: [agi] Re: Superficiality Produces Misunderstanding -
Not Good Enough
We would teach the system, PAM-P2 for example, the same
way we would teach an
infant or toddler. We would show the picture, and then say the word
"Chair" or have
the word "chair" written under the picture. We would also teach
the cognitive system
to say the word associated with the picture. We could do this for some number
of
training examples t < 25. we would then later prompt the
system with a test image,
and ask what it is, and
hopefully the system will respond "Chair". Pretty much that's
how it should happen.
The cognitive system should learn to
associate visual, auditory, proprioceptive, and
other modalities within its current, forward, and episodic models in the
same manner
as children.
~PM.
-------------------------------------------
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
