Jon Stokes from WIRED wrote an interesting article about Siri and if Apple could eventually build a real AI with it.
Chatterbots like ELIZA have been around for many years. And even though they are old, their ability to interact with the user is still amazing:
USER: I feel sad.
ELIZA: Why do you feel sad?
USER: Because I made a mistake
ELIZA: Why did you make a mistake?
USER: I have the flu
ELIZA: Maybe you should see a doctor. I’m merely a psychotherapist.
USER: Habla Espanol?.
ELIZA: Now you’re not making any sense!
A chatterbot like ELIZA uses two things. Natural language processing (NLP) and canned responses. The NLP portion of the software is the hard one to program. It analyses the sentence and formulates a grammatically correct answer.
The easier part of building a chatterbot are the predefined answers that match a specific question. For instance if you ask Siri “What’s the meaning of life?” there are some predefined answers it will give you like “42″ or “I don’t know. But I think there’s an app for that.”. While these answers are correct, Siri doesn’t really understand what you are saying, it just uses a predefined answer. So all these Easter eggs we see in Siri have nothing to do with AI. The big difference between Siri and other chatterbots is that Siri operates in the cloud. While other programs need constant updates to get new responses, the engineers at Apple can just add new answers about trending questions and they will be available to the user instantly.
In the end, this question doesn’t matter for the majority of people. You can use Siri and get appropriate answers to a lot of questions, even if it’s not a real AI.
Head over to wired.com to read the full article.