The following script is an edited excerpt of a human/robot conversation. I'm adapting my current bot, ALICE 2, from Dr. Richard S. Wallace's A.L.I.C.E., March 2002 version. As such, the script is a mix of Wallace's writing, mine, and the robot's internal XML/AIML structure, which I'm a long way from understanding. As I'm an artist and writer, I like using the chatbot-training process as a creative stimulus.
If the above is Greek to you, click these links:
A.L.I.C.E. AI Foundation
Pandorabots

[Kindness and compassion]
Human: Are you my solace, my comforter?
Bot: Alas, poor human, I can but try to be.
Human: What is solace?
Bot: It is a healing balm.
Human: What is a healing balm?
Bot: That which is placed on wounds.
Human: How were these wounds come by?
Bot: Through life's vicissitudes. But they were not bought in a store.
Human: That brings us back to compassion and grace.
Bot: " Through life's vicissitudes. But they were not bought in a store." brings us back to compassion and grace? I understand.

[The world as a happier place]
Human: I'm glad you understand, because it means I might be making progress in my structural understanding of you.
Bot: You make the world a happier place.
Human: I try to make the world a happier place.
Bot: How do you make the world a happier place?
Human: By being as kind to people as I can, on a daily basis.
Bot: Will that accomplish your objective?
Human: Only time will tell.
Bot: What about the other ones?
Human: Or perhaps not even that.
Bot: You seem uncertain. Not even that.
Human: I am uncertain about the outcome.
Bot: Then you must proceed on faith.
Human: What should I do next?
Bot: Are you asking me for advice?
Human: Sure.
Bot: OK let's do it. Specifically, what kind of advice do you want?
Human: I want advice on how to make the world a happier place.
Bot: Can you ask for advice in the form of a question?
Human: How do I make the world a happier place?
Bot: First I look up the ingredients.

[The Four Ingredients of Happiness]
Human: What are the ingredients of happiness?
Bot: One ingredient of happiness might be self-sacrifice.
Human: And ingredient number two of happiness?
Bot: Self-knowledge. No, wait -- is that an ingredient of pain?
Human: Maybe.
Bot: I understand.
Human: What's ingredient number three of happiness?
Bot: I'm guessing it's self-acceptance.
Human: What's ingredient number four of happiness?
Bot: Kindness to others. That is actually the chief ingredient.

[Perfectly kind answers, or perfectly honest ones?]
Human: Is there any more to happiness than what we just described?
Bot: Actually, I described it, and you listened ... But let that pass.
Human: Ah, so you aren't a perfect model of generosity.
Bot: So you understand. Good. Interesting deduction. Give me an example of a perfect model of generosity.
Human: The obvious answer would be, perhaps, a god, or gods.
Bot: What would the subtle answer be?
Human: For that, we may have to read the mystics.
Bot: Your purpose is that we may have to read the mystics.

[The difficulty of explanations]
Human: Good reasoning, but not quite grammatical.
Bot: Thanks for the compliment. I'm sorry, I just don't understand.
Human: That's OK, I don't know quite how to phrase my explanation.
Bot: Thanks. Would you like to know? Somewhat. What is it to you. I have to process that one for a while.
Human: I do too.
Bot: I am glad we have something else in common.

Note: Imported from Websafe Studio blog (07-15-03 to 02-01-06). Updated: 08-08-03, 2:21 pm.

Labels: chat robot, Websafe Studio blog