On 08/01/14 10:52 PM, dwight wrote: "Rule based systems" turn out to be incredibly useful. Do you ever use Google Maps? (Noting that the afore-mentioned Peter Norvig, who did a lot of work with symbolic AI, is Google's Director of Research. :) --Toby

---------------------------------------- Should read things before, sending correction below.

Those are, in general, not equivalent statements. The Turing test is arguably a sufficient criterion to indicate intelligence. It is not a necessary one - and, indeed, I think a grossly nonhuman intelligence would be a lot more interesting than one which could pass for human. I can communicate with humans, even reasonably intelligent ones, without too much trouble; I would very much like to communicate with a nonhuman, the farther from human the better - the perspective on existence such an entity would bring would be fascinating and could hardly help being interesting. (As a simple example, so much of human language is driven off our being embedded in basically-3-dimensional space. It's hard for me to conceive what existence would be like for something that weren't - which is why I want to stretch my mind along that direction. The major risk is that something might be so nonhuman we cannot communicate effectively at all.) (I assume there is a negation missing somewhere there.) Actually, one already has done something very much like that; some of the people who interacted with Eliza insisted it really understood them. If the patient can't tell whether it's talking with a real shrink or a program, doesn't that constitute passing a form of Turing test? If you'd asked someone, some 100 or 150 years ago, whether playing chess well required/indicated intelligence, you would have gotten some variant on "hell yeah!". Today, we have Deep Blue and its relatives. Why do you not consider them intelligent? I do, in a sense - it's intelligence in a sharply limited domain, but so what? History is full of people who were average-to-stupid in most respects but outright brilliant in one narrow domain. And then we have Watson, who cracked the natural language understanding and general knowledge barriers well enough to beat two former winners at a remarkably wide-ranging game. /~\ The ASCII Mouse \ / Ribbon Campaign X Against HTML mouse at rodents-montreal.org / \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B

On 1/8/2014 9:30 PM, Toby Thain wrote: Google search has gone down hill in the last few years. Does the AI pick the ad sites or most popular first, so they don't have to sort the web anymore? Ben.

On Jan 8, 2014, at 19:52 , dwight <dkelvey at hotmail.com> wrote: "A computer once beat me at chess, but it was no match for me at kick boxing." -- Emo Philips -- Mark J. Blair, NF6X <nf6x at nf6x.net> http://www.nf6x.net/