Meet ELIZA, the first chatbot that we ever got
If you are a tech enthusiast, you must be familiar with chatbots. After all, they have been quite popular in today’s tech-driven world. Especially with the advent of artificial intelligence, now more than ever, chatbots, such as ChatGPT, Siri, Google Assistant have become quite a rage. They are nothing but AI-based programmes that hold conversations though a virtual chat room. You can chat with them, ask for information and support from them, and so much more. Today, most chatbots are equipped with text message or voice-based applications and help as both personal as well as community aids.
Talking of chatbots makes us wonder which was the first chatbot ever created? Well, its name was ELIZA and she was born in the 20th century. Curious to know more? Keep reading.
Born in the hands of an MIT professor
The Massachusetts Institute of Technology (MIT) in the USA, is one of the most sought-after and reputed tech universities in the world. So, it comes as no surprise to learn that ELIZA, the first ever chatbot was the brainchild of an MIT, Professor Joseph Weizenbaum. He developed it in 1966 but as a mere caricature of human conversation and not a potential chatbot. In fact, this natural language processing computer programme was developed to demonstrate the superficiality of communication between a human and a machine. Funny, isn’t it?
Why the name ELIZA?
Not many know, but the name ELIZA was derived from the character of Eliza Doolittle from George Bernard Shaw’s coming-of-age play Pygmalion. Just like Shaw’s character was taught to speak in an upper-class accent to deceive the world, Weizenbaum's ELIZA was capable of engaging in trained conversations. But both the Elizas were incapable of understanding the nuances of language in their truest sense, during a conversation, believed the MIT professor. Hence the name. Interesting, isn’t it?
ELIZA in action
ELIZA was made in such a way that it used pattern matching and substitution methodology to simulate real-time conversation. At first, it passed on the words that the human user entered into a computer and then paired them with a list of possible scripted responses. These responses were in turn boosted by MAD-Slip, a list processing computer programming language, also created by Weizenbaum.
One of the popular scripts was DOCTOR. It was meant to imitate the well-known responses of a renowned psychotherapist named Carl Rogers. However, it was generic, without any direct medical suggestions.
Such scripts were often so powerful and human-like that they replicated the exact responses of a professional, and misleading the user with half-information. This turned out to be dangerous at times, taking ELIZA beyond a simple human caricature. She emerged as a programme that people were confiding their profound thoughts in without seeking the help of an expert.
What’s interesting here is that the so-called success of ELIZA failed to impress its maker at all though it left a mark on the world. In fact, he went on to reject the possibility of the likes of ELIZA to take over human emotions and intellect. He even argued that ELIZA was just a tool, an extension of the human mind and nothing more.
But here’s the thing. ELIZA did pioneer the future of chatbots. It was the first ever programme to successfully pass the Turing Test, the one developed by Alan Turing in the 1950s to check a machine’s ability to exhibit intelligent behaviour equivalent and indistinguishable from humans.