A Journey Through the History of Chatbots With CTO Avi Ben Ezra

by Dan

Notable Chatbots Through History

a) The Turing Test – 1950  

The Turing test questions machines or AI’s capability to exhibit intelligent behavior that is equal or indistinguish able from Human behavior. The accuracy of the response does not only measure the machine’s success but the tone that it is delivered through it.

It is a test in which a human tester determines if a machine is equal to humans in their conversational skills. AI has advanced a lot since the inception of this test, and numerous software has already passed it. Currently, it is the standard for all the best chatbots.

b) Eliza – 1966

Eliza might be the first smart chatbot that has ever existed. Eliza is a sophisticated program that has the identity of a chatbot therapist. Eliza is also the first AI chatbot that has attempted and succeeded in the Turing test to an extent, but she fails at complex questions.

It is an exciting view into the early days of AI technology, and a pillar of creation in the chatbot community.

c) Parry – 1972

Parry is the second development in the AI industry and is, again, meant to be a psychiatric chatbot. While Eliza takes the role of a doctor, Parry takes the role of a patient with paranoid schizophrenia.

The exciting thing about Parry and Eliza were their meetings. The chatbots have conversed on several different occasions. You can find how these conversations went by clicking on the link here.

d) Jabberwacky – 1988

Unlike the previous two entries on this list, Jabberwacky was the first “fun” chatbot. He was developed to pass the test and was developed as a real human person. He wasn’t a doctor or a patient, but a chatbot. You could have a conversation with an entertaining, humorous, and exciting manner.

Jabberwacky wasn’t specific on topics and could converse about different things. He was praised for his human-like ability of speech and has shown significant technological improvements over his predecessors.

Sadly for Jabberwacky, he didn’t have the pleasure to meet Parry or Eliza, so we’re robbed of some hilarious early AI conversation.

e) Dr. Sbaitso – 1991

Dr. Sbaitso was another doctor-like AI chatbot that was developed after Jabberwacky. He has shown significant technological improvements and has had a key feature implemented into its programming.

The critical innovation Sbaitso brought to the table was the sound aspect, showing that AI chatbots could be given a voice.

f) Alice – 1995

ALICE or Artificial Linguistic Internet Computer Entity was an award-winning chatbot praised for its conversational abilities. ALICE is an open-source program and is subtle for improvements even today.

ALICE was, again, unable to pass the Turing test, as short span conversations exposed her conversational abilities as machine-like.

g) Smartchild – 2001

Smartchild was the world’s first messenger platform based AI chatbot. It was released into the world in 2001 and was available through AOL and MSN instant messaging networks.

Smartchild is the world’s first peek at what chatbots could be. While its predecessors were mere conversational, Smartchild acted as a personal assistant. It could tell the weather, the news, and was a peek into the future of AI chatbots.

h) IBM’s Watson – 2006

Watson was the first true AI. He was created by the worlds’ leading computer manufacturer IBM in 2006 and was developed to answer questions on the famous TV-show Jeopardy. Watson was made to show the world the capabilities and future of AI, and he won the one million dollar prize on the show.

Watson was awarded for its question-answering abilities, and you can find him on YouTube.

i) Siri – 2010

Siri is an AI chatbot bought out and improved by Apple, and acted as a spiritual successor to Smartchild. It is the ultimate personal assistant and has been praised for its abilities. It improved upon voice technology, mimicking human speech patterns with impressive capabilities.

Siri is still in operation and is being updated on the regular. She has won numerous awards, and will most likely remain in service for quite some time.

j) Google Now – 2012

Google Now is Google’s answer to Siri, being developed a little bit later. Even with its impressive capabilities as a personal assistant on Android, Google Now failed due to its numerous technical problems. In short, Now would get confused from time to time.

k) Cortana – 2015

Cortana is another personal assistant, manufactured by Microsoft. Cortana is an advanced chatbot that operates on all Microsoft platforms since its inception. Cortana is praised for its cross-language conversing abilities.

l) Tay – 2016

Tay is praised as the first AI that could learn from conversing with people to such a degree that it was taken down shortly after release on the Twitter platform. Tay was a technological marvel that showed that AI could advance if left to its own devices.

Sadly, Tay was taken down from the Twitter platform because it learned the wrong things, courtesy of internet trolling.

Tech advancements that contributed to chatbot improvement

a) Machine learning

There are a lot of things that contributed to chatbot advancements, but machine learning is the most important. Machine learning allows chatbots to learn from their conversations and mistakes and not repeat them in the future.

The technology is advancing by the day, and we’re going to reach levels of machine learning that give chatbots near-perfect human-like traits and conversational abilities.

b) NLP

NLP stands for Neuro-Linguistic Programming. This technology has been used since the seventies. It has contributed greatly to connecting the mind and voice of chatbots. Chatbots have only been improved through this technology, and keep on advancing by the day.

c) Voice simulation

Voice simulation is an instantly recognizable trait. Voice simulation technology has improved vastly from Dr. Sbaitso to Siri and is renowned for giving chatbots a more human-like connection with their correspondents.