From: Ben Goertzel (firstname.lastname@example.org)
Date: Wed Jan 29 2003 - 07:16:19 MST
You seem to be moving in the direction of something like the
Webmind/Novamente design, which does in fact synthesize elements of neural
nets, logic (probabilistic combinatory term logic, not standard Prolog-style
logic), and a number of methods for concept formation bridging the two
What one finds when one goes in this direction -- or at least what I
found -- is that the details of the existing approaches to NN's logic, and
concept formation do NOT WORK when one tries to use them as components of an
integrated system. In practice, one has to rebuild ALL of these things from
scratch, inspired by the existing literature, in order to get them to work
as part of an integrative system.
For me, the process of doing this rebuilding took 5 years -- from 1997-2002.
The Webmind AI Engine was only a partial success in this regard -- it was
too much of an awkward combination of different things. With the Novamente
design, created in 2001, I feel we've finally "gotten it right" in the sense
of creating a detailed design that incorporates the aspects you mention but
in an elegantly unified way. But it's NOT simple -- I spent much of 2002
just trying to write up the design in a clear and complete way (while my
colleagues have worked on the C++ implementation, which is far from
complete). Now my collaborator Cassio Pennachin will spend a lot of 2003
turning my writeup into a really professional product.. and the resulting
1200+ page book will come out in 2004, by which point we'll hopefully have
50%+ of the Novamente design implemented and tested...
In short, I very much appreciate your intuition, but, you must be aware
that the step from your informal sketch to something vaguely workable is a
HUGE step with numerous difficult substeps...
I'm sure that Novamente is not the only way to flesh out this sort of
intuition either -- it's just the best way that I and my collaborators have
happened upon -- so far --
-- Ben Goertzel
> I think we are in dire need of such didactic prototypes
> if there is ever to be a *coding community* working on
> this thing. There has to be a common understanding of
> the basic concepts, and I'm not convinced that that exists
> So - there are five levels: code, modality, concept,
> thought, deliberation. What I'm going to do is just
> informally sketch something which *sounds to me* like
> it has all those levels.
> Code - no problem here, every program is made of code.
> Modality - Above all, a modality seems to be a *feature
> extractor*, operating in a specific domain. The input is
> some sort of raw data set, static or dynamic, and the
> output is a representation as a list of features, or
> objects with features. Well, neural nets can do all of
> Concept, thought, deliberation - In LOGI, these are
> described as analogous to word, sentence, stream of
> consciousness. So it seems that *propositional content*
> (think Prolog) might be enough for those top two
> levels - 'sentences' in some internal language of
> representation. The AI needs to produce descriptive
> sentences which express what its modalities tell it,
> and normative sentences which express its intentions.
> If you can get those, there is a wealth of symbolic AI
> work on putting them to use in a rational fashion.
> So, the key concept seems to be 'concept', and the way
> it bridges the gap between subsymbolic feature extraction
> and symbolic-level propositional processing. Well, one
> simple way to do that is to have concepts which are just
> lists of features. Each modality outputs an inventory
> of features possessed by its current input; the art of
> appropriate concept formation is all about picking out
> just a few of those features as 'co-relevant', worthy
> of being grouped together. And Copycat provides a model
> for *that* process.
> It looks to me like we already have all the ingredients,
> not for a seed AI, but certainly for a "general AI" as
> defined by LOGI.
> Add photos to your messages with MSN 8. Get 2 months FREE*.
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:41 MDT