Artificial Intelligence - Chatbots
...that computer can be said to be intelligent. Turing viewed language as the defining element of intelligence; he believed that by giving a machine the capacity to learn, and a willingness to ask questions, you could ‘raise’ intelligence, an entity capable of rational, engaging conversation. After computers appeared in the 1950’s, many people thought is was only a matter of time before the ‘electronic brains’ would start talking, thinking, and enslaving humanity. Predictions arose, both in science fiction and in the popular press: houses would be able to keep themselves clean, everybody would own a personal robot to complete tedious tasks and look after the kids, and computers would run the world. None of it happened. In the last half-century, billions of dollars and hundreds of thousands of hours have been spent on giving computers the ability to converse with people, but computers are still incapable of holding a proper conversation. So far, the closest thing to Alan Turing’s original vision of an intelligent computer is the chatbot. Chatbots are computer programs that are designed to converse as much like people as their programming allows. They are simple to use; you chat textually with the bot over a computer screen. The bot usually introduces itself and waits for you to respond with a statement or question. The bot consults its knowledge base or programming language and replies, continuing to do so throughout the entirety of the conversation. There are hundreds of different bots, developed for a variety of different reasons; they range from hardwired programs with simply coded patterns such as LISP, to systems built upon embedded learning algorithms that continuously expand their language knowledge base. The first well-known chatbot was Eliza, released in 1966. Eliza simulated a therapist, repeating many of the user’s questions back in a slightly different form, and recognising several keywords. For example, typing in the word “mother” would cause Eliza to respond “tell me more about your family.” Eliza shows an example of a hardwired program with simply coded patterns, her knowledge base being solely created and developed by her botmaster, or creator. The Eliza chatbot became a point of reference for other programs using similar techniques for providing a conversational interface. One program that did reference Eliza was Talk-bot, the most advanced chatbot I had come across throughout all of my chatbot research. Watch as I chat to Talk-bot. Although developed over half a century after Eliza, Talk-bot only has slightly more advanced conversational skills, yet can retain a short-term memory of the current user, enabling it to link multiple comments in order for the conversation to flow more fluently. This proves that we are still far off from being able to create a machine that can engage in a lengthy, meaningful conversation that would pass the Turing Test. Despite the many years of future research to come in order to create a satisfyingly literate machine, chatbots are still being widely developed and used for a variety of purposes. They are being created purely for fun, to be entered into competitions such as the Chatterbox Challenge, as part of interactive games, Internet information services, web site guides, e-commerce agents and more. Many bots are designed to provide specific information and direct the dialogue to specific topics. Although chatbots have mainly been used for entertainment, one company has developed a way to make a profit from such things. Conversagent designs and implements tailor-made chatbots for its clients for large sums of money as a form of advertising. Through chat applications such as AIM and ICQ, these chatbots converse with online chatters, steering conversation toward client products and allowing the chatter to ask questions and make comments. One successful chatbot campaign was the Austin Powers in Goldmember Case. New Line Cinema paid big bucks for an Austin Powers chatbot to raise awareness of the film amongst AOL chatters, previous to the film’s release. You will find this case study among your handouts for further reading. Automated Service Agents (ASAs) have also been taken abroad by large corporate companies in attempt to improve customer support, increase customer loyalty and lower customer ser...