Jump to content

Talk:Artificial Linguistic Internet Computer Entity

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia
This is an old revision of this page, as edited by SineBot (talk | contribs) at 00:06, 22 April 2009 (Signing comment by Fiskeharrison - "Turing test?: answer to question"). The present address (URL) is a permanent link to this revision, which may differ significantly from the current revision.

The Conversation of ALICE with ALICE is no longer active on that link.

IS small Brother

I was just chatting with ALICE, and somehow she got the impression my name was IS small Brother (we were discussing 1984). Anyway, I log out, log back in, and ask her who IS small Brother is. She responded "Charlie Parker." Ŏ_O --Ye Olde Luke (talk) 07:29, 15 December 2008 (UTC)[reply]


Turing test?

Is it really accurate to say ALICE can't pass a Turing test? It's fooled some people, and others not -- Turing never specified that it be 100% successful, nor that the human interrogator needed to know a lot about computers. I think ALICE *does* pass the Turing test (at least according to the initial definitions) and this shows the flaws in the idea as a test of real intelligence. If you look at something like the Loebner Prize, almost all the judges are going to be of well above-average intelligence (there's a strong self-selection effect, because high intelligence is highly correlated with being interested in AI in the first place.) To pass a Turing test, does it have to fool a highly intelligent 'computer nerd', or will fooling an average person suffice? Can we have a source saying it can't pass one? Vultur (talk) 23:10, 20 April 2009 (UTC)[reply]

It really can't. I was a judge in 2000 - and my essay on it is linked in the article, ALICE Springs - of course their are limits to the test, and the set up is artificial, but no one could mistake that for human. —Preceding unsigned comment added by Fiskeharrison (talkcontribs) 00:04, 22 April 2009 (UTC)[reply]