(My earlier posts on ChatNannies.)
Commenter Tom reports that a fellow named Andy Pryke has followed up on the original New Scientist article and met with Jim Wightman, the ChatNannie creator, and Duncan Graham-Rowe, the author of the article.
In short, when Mr. Pryke and Mr. Graham-Rowe arrived at Mr. Wightman's house he let them chat with a variant of Alicebot and claimed it was his own creation. Although Alice is one of the best chat bots around, it's nowhere near good enough to pass a Turing test. Mr. Wightman also attempted to answer some questions and to show his visitors his "about 1 million" lines of source code, but the power mysteriously went out. Darn!
The whole situation is really laughable, and I'm only posting more about it because I like to be right -- and I particularly like to rub it in when people are not only liars but jerks as well.
Whew, what a soap opera. Just spent many minutes reading the entire thread (as well as some of the linked side trails).
I agree...it is rather laughable. I'm not sure if he is simply dredging for attention or actually fishing for suckers to provide financing.