To my amazement, even though I’ve been working for an AI company (Inbenta) for a few years now, I’ve never wondered about this question before: What was the first Chatbot in history? It’s like asking oneself, Where do we come from and where we are going? pondering our existence every time we look up to the sky. If we “Google it” a little, all search results point to Eliza, the first AI chatbot.
Eliza: The First AI Chatbot
Eliza was invented in the 1960s by Joseph Wiezenbaum at the Artificial Intelligence Laboratory at the Massachusetts Institute of Technology (MIT). She/it’s considered to be the first conversational chatbot in history.
Her purpose was to show the superficiality of communication between man and machine. She did this by recognizing keywords and asking about them as if she were a psychotherapist. For example, if someone mentioned their mother in a sentence, Eliza would automatically ask them to tell her more about their family. In this way, an illusion of real understanding and interaction was created.
Eliza’s answers were so convincing that when chatting with her for the first time, some people thought they were talking to a human. However, one would soon realize that there was a limit when the user encountered implicit or subliminal messages in Eliza’s responses that didn’t really exist or didn’t have the right meaning. At this point, the conversation became incoherent. Another flaw was that she couldn’t memorize and learn from her conversations.
Current chatbots, on the other hand, have been able to overcome this obstacle and are capable of having similar interactions to those of humans. This is mainly thanks to two technologies: Artificial Intelligence and Natural Language Processing (NLP).
Who would tell Eliza that her cousins of the future would not only be able to understand and respond correctly in a useful and coherent way but also, apart from being conversational beings, they would become transactional beings. For example, activating and disabling alarms, recommending a product or service to the user, making a bank transfer or even performing a job interview.
Now we know where chatbots started, but do we know where they’re heading?
Gartner estimates that by 2020, 20 percent of the population in developed countries will use voice assistants.
Siri’s co-founder, Adam Cheyer, commented in an interview for the newspaper El País that “Soon, voice assistants will be more important than smartphones.” This suggests that virtual voice assistants could be the next technological revolution. He adds that we’re not yet aware of how important they will be in just two or three years.
It seems that the future is already here in the form of a voice.
Written by Mayra Del Canto of Inbenta.