On Tue, 15 Mar 2005 12:18:31 -0600, Paul Fidika <[EMAIL PROTECTED]> wrote: > Last I checked, Google is used by 100 million people daily, while, to the > best of my knowledge, neither Cyc nor SOAR have seen more than two or three > applications outside of the "Blocks-world", while your Novamente project, > under your own admission, is under rather tight financial straights at the > moment, so much so that you've had to start asking for donations to have the > spare time to do some "pure AGI" work...
Windows 98 is probably used by nearly as many people as google. But I don't think that that is good evidence that Windows is intelligent. I seem to remember seeing a demonstration once where Lenat asked Cyc to give examples of people who were happy, and it returned pictures of children riding bicycles for the first time or something similar (perhaps Stephen can correct me if I'm a bit off). That is probably a bit better than Google could do. As far as I know, much of Google's functionality was based upon counting the number of links leading to a page and using that a measure of its importance. > (Don't you think a 60%-complete > human-level intelligence should be capable of doing SOMETHING somone would > pay good money for?) If Cyc, SOAR, or Novamente posses even a fraction of > the "variety of intelligence-like functions" that you seem to think they do, > then where are their promised killer-apps? > A person's mind is probably a fairly complicated system. Lots of pieces depend on other pieces and lots of experience. It seems like even a small damage to some portion of it could greatly impare someone's intellectual capacities. Studies of brain damage seems to largely confirm this intuition. As I understand it, none of the authors of these projects claim to be anywhere near complete, so I don't think it fair to expect them to yield very humanlike results yet. > Eugen Leitl wrote: > >To my best knowledge (which is not much) Google currently doesn't utilize > >any advanced algorithms which could (however tenuously) be termed AI. > > If SHRDLU, ELIZA, or any of the various silly little programs Hofstadter has > ever written receive the title of "AI", why do you insist on begrudging > Google the title? I think the disagreement was over whether Google could be called "Strong AI." As in, on the path to human equivalence. This is a rather ambitious title, and none of those programs claim to deserve it. Josh ------- To unsubscribe, change your address, or temporarily deactivate your subscription, please go to http://v2.listbox.com/member/[EMAIL PROTECTED]
