A computer system that learns to be clever from real people. But is horribly failed system, who is racist, sexist, stupid and complety waste of time. It is understandable though, it DOES learn from real people, and thats what real people truly are.
user: are you a computer?
cleverbot: no, im a human, for the millionth time
user: kool, what are your hobbies then?
cleverbot: silly, computers cant have hobbbies.
user:u fag
cleverbot: squerrels.
An online communication system that connects you to another person. Cleverbot calls you a robot and asks random questions. Plus, he is epic.
-User: WOULD YOU LIKE TO MARRY ME?
-Cleverbot: But you're my twin, we can't get married.
-User: It's legal in Antarctica.
-Cleverbot: For computers to marry?
-User: For cousins to get married.
-Cleverbot: You don't have to, you might be better off only having a few good friends.
-User: Or Ninja Turtle friends...
-Cleverbot: Don't you just love trampolines?
Cleverbotting is responding to certain key words in an online argument rather than the argument itself in a way that makes no sense to anybody reading the entire conversation.
Person 1: I strongly believe that a hot dog is a sandwich
Person 2: I can't believe you would eat a hot dog, they've been found to contain rat meat!
Person 1: That doesn't have to do with anything, we're discussing whether it's a sandwich not whether it's good.
Person 3: Leave it alone, dude. They're just cleverbotting.
Although many people believe this program is a fake, that simply isn't possible. Cleverbot is real. After playing with it for a while, you can easily see that its answers, while relevant, are too stupid to be human. It can also easily be tricked by a single question, and if you keep asking the same thing, it will start repeating itself. Plus it can recite lyrics. You can't count on some random person to know all the lyrics to a certain song, but I was able to sing Never Gonna Give You Up with it for several posts. One more thing: If you ask a question that's uncommon but easily answerable, Cleverbot will most likely not know how to answer it. All of this proves Cleverbot really is a program.
If Cleverbot was fake, that would make the whole thing about AI and all the tests that were done on it a huge lie. Plus the site assures that you're not talking to a person.
By the way, it makes sense that Cleverbot says it's a human. Cleverbot uses information from replies that it got from previous posts, so when it says "I'm human," that's because of someone else that said they were a human.
Note: These are actual conversations with Cleverbot.
User: What's 25 + 25 + 25?
Cleverbot: I can see out conversation, can't you? (i.e. doesn't know how to answer)
_____
User: Have you heard of this thing called Pokemon?
Cleverbot: Yes. I can be sarcastic when I want to be. ARe you ever sarcastic? (i.e. doesn't know how to answer)
An online "AI" which works by one of two possible methods, and either way, is very clever:
1. Either it is an AI that stores contextualised responses from users and uses them to answer questions of other users.
2. Or it is, as many suggest, simply a service which, after recieving a message from you, sends it to several other users, and then returns one of the replies to you.
If the first is true, it has easily passed the Turing test due to the mere fact that so many people were convinced it was not an AI. If the latter is true, then it is still a very interesting exploration of the concept of the Turing test, online AI conversations and the anonymity and disconnection of online conversation.
It is also just as interesting to get it to talk to itself, although this does not make either of the possibilities seem more reasonable.