What is human?

In “What’s It Mean to Be Human, Anyway?,” Charles Platt describes the Alan Turring’s idea that if a computer can successfully imitate a human being during what basically amounts to a conversation, it should be practically considered a human. This, of course, was in 1950 and, judging by old episodes of The Twilight Zone, people were a bit more hopeful about the capabilities of computers back then. And, of course, no computer has come close to matching the human characteristics of Asimov’s Multivac. However, in an apparent attempt to encourage humanistic robot technology, Hugh Loebner, of inflatable disco floor fame, began hosting a competition for programmers to make their machines imitate humans. (Maybe he wants a technological utopia, maybe he wants sexbots. No one knows for sure.)
The participants seemed to have a common strategy: have enough preprogrammed responses to prepare for every potential question. No analysis, no thought, just a set answer based on a few textual cues. I wouldn’t quite call that imitating human life. And many of the critics of Loebner’s contest agree with me. A researcher mentioned in the article claimed his contest favored cheap tricks over the development of real artificial intelligence.
It would appear that the Turring test is flawed; these machines don’t imitate anything. They just spit out answers to what their programmers thought were likely questions. Important to note is that the winner wasn’t designed for this competition at all; it was meant to be an encyclopedia style computer for dispensing sex advice to shy Canadians. That’s all these machines are- user-friendly encyclopedias. They don’t imitate life, they store and dispense our knowledge. We’ve all, at least once, said “excuse me” after bumping into a store mannequin. We get confused and think it’s a human. It tricks us. Would we ever suggest that, for all practical purposes, we consider it to be a human being? Similarly, would anyone for a minute suggest that someone severely catatonic is not a human being?

If the Turring test is a flawed way to determine whether or not a computer is worthy of consideration as a human being, by what method can we determine the humanness of a computer? For that matter, what do we use to determine the humanness of a human?