On 11/02/2008, Richard Loosemore <[EMAIL PROTECTED]> wrote: > As far as I can tell, any scenario that anyone has imagined for a > sub-human-intelligence robot that would be useful enough to sell in the > market place contains limitations that will doom it in the same way. > That last 2% will always be enough to doom it.
This is a good counter argument. What you're suggesting I think is that there may be no easily commercialisable and incremental path between the kind of technology we have today and machines which are at least as useful as humans can be. There is something to be said for this, and the fact that progress in robotics has been relatively slow is precisely because there are a few key problems for which partial solutions are just not viable. > But now, by contrast, if you are assuming (as Matt does, I believe) that > somehow a cluster of sub-intelligent specialists across the net will > gradually increase in intelligence until their sum total amounts to a > full AI, then you are begging some enormous questions. The "army of experts" is only one possibility. Probably like most people on this list I think producing more intelligent machines is going to require a more closely integrated cognitive architecture. Integration however doesn't mean that the system has to reside on a single computer or physical device. ----- 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/?member_id=8660244&id_secret=94603346-a08d2f
