Cleverbot apparently passes the Turing test. I think this says more about the Turing test than it does Cleverbot.
I suspect that Alan Turing’s idea of the Turing test was a little different to what is commonly implemented. Ultimately the concept of having a machine convince you that it is human is a pretty sound idea, and no alternative immediately presents itself. But I’m sure that Turing didn’t think a bunch of idiots casually chatting to a bot via a text-link for five minutes would be a good indicator of that. Even if he did, he certainly didn’t know that the internet and instant messenger would greatly lower the bar for what we think is reasonable human conversation in such a setting. It is fairly difficult to very quickly distinguish between a bot and a human on instant messenger; humans don’t always respond well to being tested and asked to do strange things, and casual conversation is pretty loose; if the bot is driving it then it can quite often ignore your response and go onto its next statement, and it still feels like a natural conversation for a few messages even if it’s entirely scripted (trust me, I’ve talked to a bot for a minute or two before I realised it was about to try to get me to visit a website, and it was only when it came out with the URL that I was sure).
There are other forms of communication than IM. Cleverbot cannot write an article, it cannot write an essay, it cannot answer short ‘discussion’ questions, it can’t tell me in 200 words or fewer whether or not it considers the Turing test to be useful and why. Ultimately, if a piece of AI is able to ‘think’ in any meaningful sense of the word, and if the output of its thought is written language, then it should be able to do these things. If it can have a reasonably convincing one line back-and-forth conversation but it cannot write a short paragraph on a single subject, then whatever it is doing, it is clearly not ‘thinking’.
me: What colour was George Washington’s favourite white horse?
Cleverbot: White. Duh, you said so yourself.
me: What colour was George Washington’s favourite magenta horse?
Cleverbot: I don’t know. The song doesn’t include that information.
Cleverbot: Who is your favorite pop artist?
me: Lady gaga.
Cleverbot: Lady gaga is not a philosopher.
me: Um I know.
Cleverbot: What about Lady Gaga, then?
Cleverbot: Can computers think?
me: In theory yes. But you can’t.
Cleverbot: I just did.
me: No you didn’t.
Cleverbot: Yeah, I sent her the message. She won’t get it until tomorrow.
me: What does it say?
Cleverbot: The weather doesn’t say anything.