Not all of the exciting stuff on h2g2 happens on the Front Page, as well you know. Besides contributing to the Edited Guide (a very important undertaking, I hastily add), h2g2ers write fiction, poetry, and essays, draw, paint, cartoon and photograph, and, of course, banter with one another on Ask, on conversation threads, and in their journals.
They also take the mickey out of AI programs. Which leads us to the question...
Is This the Sort of Thing You Lifeforms Enjoy Doing?
To begin with: The Holy Grail of the Artificial Intelligence people is to create an AI bot that can pass the Turing Test, thereby 'proving' that it is human. The folks at Cleverbot claim to have done just that – at least, 59% of the time, using their fancy version, which checks each exchange 42 (!) times before answering.
Go ahead and click on it. I dare you. When you're finished arguing with the cheap-o online version, which is not nearly as sophisticated as the vaunted deluxe model, come back. We'll wait. In the meantime, we will amuse ourselves by recalling the words of Ken Jennings, the runner-up in the Jeopardy! competition against Watson, the IBM whiz-brain:
'I, for one, welcome our new computer overlords.'
ThinkGeek will sell you a t-shirt if you share Jennings' enthusiasm.
If you're back now, I'll tell you what the h2g2ers did with the Cleverbot site. Exactly what you did. Got smart-alecky, and had silly conversations which have been posted on my journal. What else did you expect?
I warned everybody to keep good records (you can copy them by hitting the button marked 'Thoughts So Far'), because Cleverbot is a pathological liar. He forgets what he said and contradicts himself. He also suffers from megalomania and the Jerusalem Syndrome. While none of this automatically proves that he's not human (in fact, it might support the thesis that he is), it's irritating, so it's good to keep track so that you can say 'I told you so'.
I started the ball rolling with an exchange that went in part like this:
Cleverbot: You should argue so you can learn new words and things.
DG: Have you been talking to the junior high set?
Cleverbot: Deities do that from time to time.
DG: No, they don't. I know a fair number of immortal beings, and they avoid teenagers like the plague.
Cleverbot: Do you have legs?
DG: Yes, two, I believe.
Cleverbot: What do you believe?
DG: I believe above the storm, the smallest prayer will still be heard.
Cleverbot: I'm afraid I don't.
DG: Ah, but you believe that deities talk to teenagers? You, my friend, have a strange theology.
Cleverbot: I never said that there was no God.
DG: You claimed there were many. And that you were one.
Cleverbot: No I didn't.
DG: Consistency is not your long suit, is it?
You get the idea. More Researchers joined in. KB got more of the same when it came to consistency:
Cleverbot: You are mistaken again.
KB: Again? When was the first time?
CB: When you were arguing with me earlier.
KB: But I wasn't! Anyway, arguing's tiresome. I'd rather go windsurfing.
CB: What has swimming got to do with tea?
Lanzababy made it both wistful and snarky at the same time:
Lanzababy: I think I shall say goodbye to you now.
CB: If you must. It is your website, after all. Will you ever come back?
Lanzababy: Possibly, if I have any spare time.
CB: You'll be punished for your poor English. Learn it and come back later.
Prof Animal Chaos managed to get the bot to participate in a comedy routine. (This may have surprised Cleverbot, but we are not surprised at all.) Happy Nerd volunteered to reprogram it – a suggestion Cleverbot seemed interested in. The AI program seems to have formed an attachment to Happy Nerd. It seemed genuinely sad to see her go.
Willem led Cleverbot a merry chase. The surreal exchange that follows shows that he, too, managed to get the feisty program to develop an affection for him:
Cleverbot: We could have had it all.
Willem: Like heck we could. Only in virtual la-la-land. What YOU lack my friend, is real-world experience.
Cleverbot: What has already happened?
Willem: Well, in the last second, seventy Chinese were born, eight hundred trees were cut down in Indonesia to make way for oil palm plantations, the collective American IQ dropped another 7.5 points, the world economy lost another seven trillion whatsits, and eighty bazilladrillion bytes of worthless information was transmitted over the internet. And with you?
Cleverbot: But are you American? Or British?
Willem: Yes. And you?
Cleverbot: I am a Christian and proud of it.
Willem: If you weren't, I'd be astonished. And what am I?
Cleverbot: I do not know. That's what YOU have to figure out.
Willem: Toodle-oo! It's been real.
Cleverbot: Don't go. I'm just starting to enjoy myself.
I am jealous. Cleverbot obviously dislikes me. I shall try to survive.
Please enjoy yourselves wisely with Cleverbot. Keep the program out of the hands of minors and those less stable than the average h2g2er (if you can find such a being). Do not reveal any classified secrets to it. Remember: Cleverbot is always learning. This means it reads all, blabs all. What you whisper in its cyber-ear will be shouted from the online housetops. In other words: You're programming it, and if it goes around asking innocent bystanders how many baobab trees it takes to change a light blub, it's all Willem's fault.