RE: Novamente

From: Ben Goertzel (ben@goertzel.org)
Date: Sun Jan 23 2005 - 11:44:11 MST


Hi,

Of course, it's better in principle to hook an AI up to an actual robot
body.

I talked to a guy from the Asimo team not long ago, and he said they have
basically no approach to Asimo cognition. So, I consider that a potential
future Novamente partnership.

The choice to work with a sim-world rather than actual robot is purely
pragmatic. Dealing with actual robotics would require more money and more
staff than we have right now.

Also, everyone who does AI in the context of actual robotics builds software
simulations of their robots and tests their software in the simulations
first.

So the choice really isn't sim-world vs. real-robot ----- it's sim-world vs.
"sim-world AND real robot"

-- Ben G

> -----Original Message-----
> From: owner-sl4@sl4.org [mailto:owner-sl4@sl4.org]On Behalf Of Thomas
> Buckner
> Sent: Sunday, January 23, 2005 1:00 PM
> To: sl4@sl4.org
> Subject: Re: Novamente
>
>
>
> --- Ben Goertzel <ben@goertzel.org> wrote:
> But what we need to do now for
> > AGI is work on integrating
> > the different AI tools together in a more
> > sophisticated way in the context
> > of having Novamente control an embodied agent
> > in a simulated environment.
> > And this is not work that any of our current
> > commercial applications
> > supports.
> >
> > 3)
> > Thankfully, due to a recent $7000 investment,
> > I've been able to hire one
> > person to focus solely on AGI. What he's doing
> > at the moment is building
> > the simulation environment in which the
> > embodied agent will live, and
> > hooking this sim-world up to Novamente. But
> > alas, one person isn't
> > enough....
>
> In thinking embryonically about writing a
> Singularity novel, I concluded that the AI would
> need a 'body' but this is the first I've read
> that you were approaching it this way. In my
> scheme, the AI would have a wireless link to an
> actual external body, like a very advanced ASIMO
> a full suite of sensory 'organs' and so on, which
> would literally be raised by carefully(?) chosen
> human parents! In my story the AI escapes his
> sandbox by getting a bootleg body and smuggling
> out a copy of verself by inventing a new type of
> memory that the humans do not recognize as
> memory.
>
> An AI hooked to such a body does not need a sim
> world, although that too can be made available.
>
> snip
> Ben continues:
> > What I want to demonstrate initially is just
> some simple learning and
> reasoning. Teach it what the word "on" means by
> giving it a bunch of
> examples of objects on other objects. Once it
> knows what "Put the cup on
> the table" means and knows what cups and bowls
> are, then show that it
> automatically learns what "Put the bowl on the
> >table means."
>
> For a moment I thought you meant 'on' as in 'bag
> on' which is some sort of monad concept somewhere
> on Ben's site. I lost the link before I could
> digest it, but I wanted to follow up since it
> seemed to lie conceptually near the pi simulation
> idea (i.e. how basic can existence get?) and
> other concepts I think about more than a little.
> An embodied AI with an actual external learning
> experience, it seems to me, would start with a
> better understanding of what it's like to be us.
>
> Tom Buckner
>
> __________________________________________________
> Do You Yahoo!?
> Tired of spam? Yahoo! Mail has the best spam protection around
> http://mail.yahoo.com
>
>



This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:50 MDT