Share Site Map Contact us Home Page
Home > Virtual Personalities > Alan > Selected Alan Logs > Session 47000 Jan. 21, ...

Selected Alan Logs

Alan logs
Selected Alan Logs
Alan logs with other bots
Best of Alan
Private Virtual Personalities
Available PVPs


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?
U> yes
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?
U> yes
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.
U> yes
A> Ok. Shoot.
U> knock knock
A> Who's there?
U> ur
A> you are who?
U> urainass
A> Hillarious. Lol.
U> l spock to you last night
A> I think it's time for a new subject, Jon.
U> space
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> o
A> Oh.
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?
U> yes
A> Basically, chatterbots are a logic system of 'if/then' statements. They look for keywords, and then go to the answer that fits best.
U> oh
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.
U> oh
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.
U> oh
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?
U> books.
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.
U> p
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

  Printable version
Ai Discussion Boards  Talk about it