On Fri, Apr 5, 2013 at 8:27 AM, Ben Goertzel <[email protected]> wrote:

> > Specifically, how would you optimize Google's cat face
> > recognizer
>


> But I'm not going to try to pack the details of my AGI thinking into an
> email...
>

Me neither. Unless you send a truckload of a dollars to my Virgin Islands
anonymous account. But if you bring together some of the things we know
about sub-human classifiers and human cognitive performance then you might
conclude that the general problem (our remarkable potential mastery of
visual scenes, audio scenes, tightrope walkig and even wine tasting)
probably is prediction/generation/imagination related, we are just too good
generating possible and impossible body-head-limb positions for pretty
much all mammals, while more abstract models stretch our abilities enough
to be part of IQ tests, nobody has included cat pics in IQ tests yet,
right? We need to bite the bullet and generate scenes automatically,
compare them to input, and accept the best match not only as a cat or two,
but as the understanding/explanation of the scene too - whether audio,
video or whatever.

AT



-------------------------------------------
AGI
Archives: https://www.listbox.com/member/archive/303/=now
RSS Feed: https://www.listbox.com/member/archive/rss/303/21088071-f452e424
Modify Your Subscription: 
https://www.listbox.com/member/?member_id=21088071&id_secret=21088071-58d57657
Powered by Listbox: http://www.listbox.com

Reply via email to