From: Joshua Fox <[EMAIL PROTECTED]>
>To those doing AGI development: If, at the end of the development stage of your project -- say, after approximately five >years -- you find that it has failed technically to the point that it is not salvageable, what do you think is most likely to have >caused it? Let's exclude financial and management considerations from this discussion; and let's take for granted that a failure
>is just a learning opportunity for the next step.
>is just a learning opportunity for the next step.
As far as I know, every AGI project in the last 50 years has either failed or is still ongoing (e.g. Cyc). I think the common reason is
that the designers underestimated the difficulty of the problem. AGI requires vast knowledge and a lot of computational power. I don't think there are any shortcuts, or we would have found them by now. Moore's law will solve the computational bottleneck, but I think to obtain all the required knowledge and experience that makes us human, you need to duplicate the human body.
I can't answer for myself, since my project ( http://cs.fit.edu/~mmahoney/compression/text.html ) is still ongoing after 7 years. But I am not trying to solve AGI, just the more narrow problem of language modeling, which would be AI in the sense of the Turing test. For the closely related problem of text prediction, adequate training data is readily available.
-- Matt Mahoney, [EMAIL PROTECTED]I can't answer for myself, since my project ( http://cs.fit.edu/~mmahoney/compression/text.html ) is still ongoing after 7 years. But I am not trying to solve AGI, just the more narrow problem of language modeling, which would be AI in the sense of the Turing test. For the closely related problem of text prediction, adequate training data is readily available.
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]
