Well heres a little excerpt from one of the conversions where I was testing the cleverbot and it was getting a little weird.
Cleverbot: Now, you can ask me any questions you want.
User: You know the rules.
Cleverbot: What are the rules?
User: Cat I'm a kitty cat.
Cleverbot: Thank you for proving my point.
User: What was your point?
Cleverbot: That this conversation is odd because you're only simulating a human being.
User: Well, what else should I be stimulating?
Cleverbot: Ha Ha funny. You should say hi or something.
User: Ok hi.
Cleverbot: Okay? That's all? No apology for being rude?
User: Ah I though you meant stimulate and not simulate.
Cleverbot: But even if you are human, you are not 'being' human, therefore you are negating any humanity that you have at the moment and are merely
an extension of communication.
User: If I'm human but not being human then what am I being?
Cleverbot: How do I know I am 'speaking' with a human.
User: Well, what else would you be speaking to?
Cleverbot: Possibly a computer. Possibly a person. Can a computer be good? Can a person?
User: But you are able to tell the difference between a person and a computer?
Cleverbot: I'm not in you, and I cannot say whether or not I am a computer. That info. Is claaified.
User: So you have to 'be in me' in order to know if I'm a computer or a person?
Cleverbot: I'm a person.
Anyways now it's clear that cleverbot is fooling alot of people into it actually being human, just not in a respectful way. This may just be a mix of
having it's responses being to nonsensical and it doing the "you are cleverbot" talk that it has learned. From what I see this is just a very basic
demonstration of what some software developers want an aspect of robotics to evolve into(not including the nonsensical part...). This bot is gaining
information from everyone who talks to it so of course it's going to not make sense at a lot of points. This can make some people believe it's
another human if they have certain experiences on the internet that give them this gut instinct when they are messaging someone else. These feelings
are wrong however and can easily be confirmed by just doing some tests to see how the bot interacts and learns.
The cleverbot is very random, some people may just have the strict conception that bots have to be programmed to be random in certain ways. This may
comfort some people but the truth tha this software conception is outdated. I'll expand on that later if I feel like it. Excerpt
[edit on 9-11-2009 by xicor]