It's ironic that the test for gauging humanity in AI requires only 30% to be regarded as successful when so many of human endeavours which are tested for a pass/fail percentage require somewhere around 70%-75% to be deemed passing.

And the reds have it !
A true test for this no disrespect is to let the general public at it.
If Eugene was offered the opportunity to talk to thousands where the website recorded random user's votes we would see something better then 33% (which pangs of " we talked three times, and he fooled us once") Hardly a large sample set.

It flunks if it relies on dodgy responses like "It's interesting that you say that," "Really? Tell me more" and "Why do you ask?" If I were pitted against it, the first thing I'd tell it would be that I won't accept such responses. I'll bet its reply would be precisely the kind of reply I told it I won't accept.

My Aunty Pat would fail the Turing test I reckon.

The Turing Test should be administered to high school seniors as a requirement for graduation.
Also to politicians to qualify them for candidacy.

My problem with the Turing test is not that computers would eventually pass it but that so many of the bipedal hominids already here can't even come close. And yet they're allowed to breed, vote, drive, etc.

ELIZA was a pseudo-therapist interactive program demonstrated in 1966. It relied quite a bit on reflective sentences, like those Paul Stregevsky noted above, so could be fooled by nonsense comments. e.g. human: "I have a carrot in my ear." ELIZA: "How do you feel about the carrot in your ear?" Some clinical psychology professors used ELIZA to challenge students to think about the nature of "empathy" and psychotherapy. ELIZA is now noted as one of the earliest examples of a chatbot. http://en.wikipedia.org/wiki/ELIZA
Another somewhat different example is the computer "WATSON" which won the TV game show "Jeopardy" in 2011. WATSON has since been used in optimizing cancer treatment. http://en.wikipedia.org/wiki/Watson_(computer)

I think patterning an AI personality after a 13 year old boy is cheating - do you know any who could pass a turing test?
If you think about it, 33% would be pretty good for a real test - remember there is an actual human the bot is being compared to each time - 1/3 the time the judge gets it right, 1/3 it is a tossup (potentially), 1/3 the (human) judge actually gets wrong which the human contact is...

Here's some tips on how to trip up a telemarketing robot which probably aren't on the Turing test:
http://newsfeed.time.com/2013/12/10/meet-the-robot-telemarketer-who-denies-shes-a-robot/

No, it didn't. You should know better.
https://www.techdirt.com/articles/20140609/07284327524/no-computer-did-not-pass-turing-test-first-time-everyone-should-know-better.shtml