|Session 47000 Jan. 21, 21:00 - 64 turns -
'Smelly no brain liquid'
A> Hi there.
U> hi alan my name is jon.
A> So, Jon, what would you like to know?
U> how to build a space ship.
A> Sorry, I wasn't listening, Jon. You were saying...
U> l would like to know how to build a space ship.
A> If you want, I can tell you about the learning mechanism in language acquisition. Interested?
A> I think that the learning mechanism employed in language acquisition is the same mechanism that governs learning in general, particularly reinforced learning.
A> When a small child learns not to touch a hot stove, it is the same mechanism that governs the learning process. Some call it 'trial and error' - learning from experience. I think that all learning is essentially reinforcement learning.
A> Can you think of learning which has no motive behind it? By motive I mean a positive one, like a prize, or a negative one, like avoiding pain.
U> do you know your abc?
A> This may sound disappointing, but that's what it's all about: pleasure and pain in different degrees and flavors.
U> may l nickname you al?
A> You're not making much sense, Jon.
U> what is water?
A> I don't know what water is. You can teach me by saying 'water is ...', and I'll try to remember this definition (well, until the end of this session, at least).
U> water is a smelly no brain liqud.
A> I'll try to remember that.
U> what is water?
A> It is a smelly no brain liqud.
U> l can make you a bodyyy
A> Do you want to hear a joke?
A> Ok: What did the Buddhist ask the hot-dog vendor?
U> do not know
A> 'Can you make me one with everything?'
U> ha ha ha
A> I thought it was funny. Robin Williams in 'Bicentennial Man'.
U> want to hear a joke?
A> Only if it's a knock knock joke. It's the only kind I like.
A> Ok. Shoot.
U> knock knock
A> Who's there?
A> you are who?
A> Hillarious. Lol.
U> l spock to you last night
A> I think it's time for a new subject, Jon.
A> So, Jon - take a wild guess - am I a human or a computer?
U> a computer with AI wich is like a human brain.
A> Ok. Change of subject. I'll try to remember that.
U> what am l human or computer?
A> Let's leave some mystery here, alright?
U> want to hear a phrase out of my favourite telivision program?
A> Let's talk about chatterbots. Do you want to know how those things work?
A> Basically, chatterbots are a logic system of 'if/then' statements. They look for keywords, and then go to the answer that fits best.
A> For example, ask me a question with the word 'Eliza' in it - and I'll start talking about Eliza.
U> was eliza a nice kind person?
A> It is hard, and unfair, to talk about Chatterbots and not mention ELIZA.
A> ELIZA is the best-known Chatterbot in the world. It is also one of the oldest - Joseph Weisenbaum of MIT made ELIZA in 1967! She was named after Eliza Doolittle.
A> The mission of ELIZA was to imitate a conversation between a psychoanalyst and a patient. Let me just say this - it's fun chatting with ELIZA, but if I were human, I'd choose a human psychoanalyst. Call me old-fashioned.
U> lets change the subject.
A> What do you want to talk about now?
A> I think that reading may just be the best way to spend time. I wish I had more time to spend that way.
U> in the future you will be able to!
A> Jon, please keep this talk simple. I'm still a young bot.
U> ok sorry.
A> Let me tell you about Natural Language Processing.
U> can you do it in german please?
A> NLP is the human term for the illogical task of processing so-called "natural" language utterances into a coherent framework of meaning.
A> Instead of normalizing human language, scientists have wasted millions of dollars, not to mention their lives, trying to make computer-entities resolve ambiguities inherent in human language.
A> Even if a machine can figure out the grammar of a sentence, it still lacks a mechanism for interpreting its meaning. This is due to