24 November 2013

Artificial Intelligence: The Future of Chatterbots

An automated online assistant uses AI to help users.

Imagine that you are calling the phone line of a major business.  More often than not, you will encounter an interactive voice responder.  Your experience with the voice responder may be a familiar one—a recording guides you throughout a menu and slowly reads out each option.  You press the button corresponding to the desired option, and the recording reads out the next menu, ad infinitum.  Many people find this process to be slow and tedious.  A lot of this perceived tedium comes from the fact that such menus are counter-intuitive to natural human communication; it is in our nature to communicate specific ideas in short, easily-understandable phrases.

In recent years, however, computer scientists have been looking into systems that can recognize and respond to natural human speech, albeit to a limited extent.  Such a system would be a product of artificial intelligence, a branch of computer science that studies how machines can process and respond to input.  Some current computer programs that recognize human language include chatterbots, software that simulates a conversation with a user, primarily for entertainment purposes.  One of the most popular chatterbots is Cleverbot, who constantly develops speech mannerisms from interacting with humans[1].  Some broadcasters on websites such as Youtube record themselves interacting with Cleverbot and other similar bots, often to humorous effect[2].


Using chatterbots for professional use is still limited, however.  Some chatterbot technologies such as SitePal[3] and AlterEgos[4] have been made for use by businesses, but they are rarely seen on other websites.  Perhaps good website layout can make such chatterbots unnecessary.  Still, possibly in the future, when services become used by more people and technical support staffs become overworked, artificial intelligence software will be able to supplement human assistance.