On Monday 25 September 2006 16:48, Ben Goertzel wrote: > My own view is that symbol grounding is not a waste of time ... but, > **exclusive reliance** on symbol grounding is a waste of time.
It's certainly not a waste of time in the general sense, especially if you're going to be building a robot! But I just don't think it's on the critical path. > Novamente utilizes a combination of grounding of symbols in > simulated-embodied experience with ingestion of information from > existing databases. I believe this sort of combination is optimal, > rather than purely relying on data sources with no attention to > embodied experience.... You're definitely right if the databases you're thinking of are things like CYC or Wikipedia. General English text is shot thru with dependencies on physical/body understanding. But it's also shot thru with social/emotional dependencies which probably amount to 10 times as much primitive semantics. > My own view is that all serious learning algorithms are inevitably > going to scale exponentially -- so the whole art of AGI design is in > figuring out appropriate tricks for making the exponent and the > constant outside the exponential function "not too large" for problem > classes of practical import... > Well, in some sense, Solomonoff/Levin search is general, but of course it blows up like a balloon. People like John McCarthy (and of course Solomonoff) talked about this kind of thing as far back as the '50s. EVERY attempt at a general learning algorithm has the same problem: it's exponential in the complexity of the structure created. I don't think that just reducing the exponent makes a general learner. I think that what has to happen is that the system itself has to be able to learn the kind of things that can reduce the exponent -- so it can get more efficient as it goes. At any given point it looks exponential, but it somehow keeps finding "just one more trick" that lets it build ever-more sophisticated structures. Josh ----- This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/[EMAIL PROTECTED]
