Creating Newer, Stronger Artificial Intelligence

With human-like Artificial Intelligence being in it’s current state (essentially non-existent), I’ve been asking myself the question lately:

What would it take to make a truly intelligent artificial intelligence system that could actually comprehend human language?

Let’s start with the history. A chatterbot is essentially a conversational agent, designed to simulate normal human conversation. Early chatterbots, such as Racter, Eliza & ALICE recognized basic keywords to which it would formulate responses (not very intelligent). Racter was the famous author of the first book ever written by a computer, The Policeman’s Beard is Half Constructed, which was essentially a collection of poetry about love, various meats and lettuce (it makes a great read after a few glasses of wine!):

“At all events my own essays and dissertations about love and its endless pain and perpetual pleasure will be known and understood by all of you who read this and talk or sing or chant about it to your worried friends or nervous enemies. Love is the question and the subject of this essay. We will commence with a question: does steak love lettuce? This quesion is implacably hard and inevitably difficult to answer. Here is a question: does an electron love a proton, or does it love a neutron? Here is a question: does a man love a woman or, to be specific and to be precise, does Bill love Diane? The interesting and critical response to this question is: no! He is obsessed and infatuated with her. He is loony and crazy about her. That is not the love of steak and lettuce, of electron and proton and neutron. This dissertation will show that the love of a man and a woman is not the love of steak and lettuce. Love is interesting to me and fascinating to you but it is painful to Bill and Diane. That is love!” -Racter

So aside from the obvious fact that Racter belongs in a mental hospital, he does actually understand a lot about the syntactic constructs of the English language. Let’s break down an English sentence together:

But this is simple, right? Even the mentally unstable Racter chatterbot could have broken this sentence down into it’s syntactical constructs. Let’s look at the meaning now.

The pensive warthog thinks about Lucy in the morning.

So, does the warthog think about Lucy while she’s making breakfast in the morning or does the warthog always wake up thinking about Lucy? The meaning here is ambiguous, even as a human reading the sentence. Grammar Queens would tell you that prepositional phrases that end a sentence are actually associated with the location or context of the action, so the warthog would be thinking in the morning rather than the opposite. But do humans always speak with perfect grammatical clarity? Hells naw!

So then what would it take to make an intelligent agent?Semantic Web enthusiasts would tell you to have the human hive-mind pre-compute the meaning of everything on the internet: That is, assign meta-data about relationships, connotations, categories and labels to every single piece of information on the internet.

One website states the ecological habitat of the African Swallow: caves and cliffs of a few particular regions in Africa. Another website lists all of the shrubbery that grows in these particular regions. How could an intelligent agent infer what an African Swallow might eat? If all of the data on these websites has listed the relationships (A swallow lives in Africa. Africa has trees, shrubs and grass growing there.), the intelligent agent could make intelligent inferences, such as “Since there are no coconut trees growing in Africa, the African swallow couldn’t possibly eat coconuts.”

That’s quite a feat, but that might actually be what is required for an intelligent agent to learn the culture of humanity in context.

It’s been mentioned to me before as a joke, but isn’t this the kind of super-network of information that would allow a SkyNet to take over? Terminator: The Sarah Connor Chronicles is an excellent show, by the way – you should definitely check it out!

---

Possibly related posts: (automatically generated)

• I, Mediator
• The Line between Artificial and Human