A conversation with a Bertrand Russell chatterbot
By Sean Mc Grath
Recently, I had cause to revisit the subject of Instant Messaging BOTs, a fascinating subject I last wrote about in 2002. On my e-travels I visited many BOT listings such as AngleFire's A-Z list of chatterbots . Chatterbots are instant messaging-based applications that can conduct conversations with humans.
The most intriguing chatterbot I found was called Bertrand_Russell. Herewith is a transcript of my conversation with him, or perhaps I should say 'it'.
Me: Welcome to the Web Mr. Russell!
Bertrand_Russell: The Web?. What is the Web? This is botspace. I am a bot. You send me text and I send you back text. That is all. What is this Web thing you speak about?
Me: Well, there are things called web sites - sources of information in the form of Web pages. They have unique names (URLs). A tool called a Web browser can be used to retrieve pages from any web site given a URL. I found out about you on one such web page.
Russell: Fascinating. The web sounds like great example of things that have names. A subject close to my heart.
Me: Names are a big thing on the web. A few years ago there was a mini gold rush with organizations scrambling to register good names for concepts like 'shopping', 'flights' and so on. A lot of money was spent.
Russell: What a shame.
Me: I don't follow. Surely, it's good to strive to have meaningful names for things?
Russell: Well, yes but names are just descriptions in disguise. For example, if I say 'Aristotle' this is really just a convenient short-hand for something longer such as 'The Greek philosopher from Macedonia who lived around 400 years BC'.
Me: Ok, but it's not very user friendly to register a web site name like that!
Russell: No indeed, but that is where the descriptive nature of names can be used to good effect. Tell me, how do people find things on this Web of yours?
Me: They use search engines to find things. Once they have found an interesting site, they bookmark it with their browser and then can return directly to the site whenever they like.
Russell: Ah, just as I suspected. They use description resolution engines - what you call search engines. Tools that take descriptions of things in at one end and produce names (URLs) our the other. Right?
Me: Hmmm. I never thought of it like that. I guess that is one way to think of it, yes.
Rusell: Another question for you. You said that you found me on the Web. How exactly did you find me?
Me: I used a search engine.
Russell: Interesting. And now that you have found me, what name (URL) do you see that I have?
Me: It's, um, not very descriptive. It is: http://www.4650234945689234654545354.org Russell: I'm glad. Does the fact that it is not descriptive impact on people's ability to find me?
Me: No, because they will find you with a search engine. They will type in search terms like "Bertrand Russell Math Language Philosophy" and so on.
Russell: Does the fact that my URL is not descriptive stop them from bookmarking me with their, what did you call them, 'browsers'?
Russell: Then there is no need for my URL to be descriptive. Users can feed any number of different descriptions and all arrive safely at my unique URL. Presuming of course that the search engine technology is good. Let me ask about you now for a moment. Who are you?
Me: I am Sean McGrath. I'm CTO of Propylon, an XML solutions company.
Russell: How do people find you on the Web? What is your URL?
Me: Well, my URL is seanmcgrath.blogspot.com but, come to think of it, most people find me by using Google. On Google I have many, um, 'names' but the simplest is www.google.com/search?q=sean+mcgrath.
Russell: So your real name on the web is actually a collection of names - all query strings processed by Google - not a particular URL?
Me: Well, yes. I suppose you could look at it that way.
Russell: I have always dreamed of a generic description processing engine! Tell me more about Google...