From: Ben Goertzel (ben@goertzel.org)
Date: Fri Dec 13 2002 - 15:08:07 MST
> > Bill, I think that computationally-plausible temporal credit
> assignment is a
> > bigger can of worms than you're realizing ;-)
>
> Well, in my book I say it will be 100 years before we can solve it
> (including having machines with sufficient computing power). Do you
> think its a bigger can of worms than that?
>
> Cheers,
> Bill
Nah... as you know I think I have sketched a software solution to this
problem already, and that we can have it implemented & tested and learning
away in 2-10 years depending on funding...
So I guess the issue is NOT that I think the problem is harder than you do
;) I guess I would have liked you to say more in your book about the reason
*why* it's hard, and strategies that might be taken to solve it.
But then, it's only one brief book and that wasn't its aim.... And if you'd
said all that in there, there'd be less left for the Novamente book when
it's finally done ;)
ben
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:41 MDT