Blog: History of the Chatbots
Since the 60’s we have evolved from a paranoid schizophrenic chatbot to Siri, Google Now, Cortana or Alexa, among others…
In recent years, the word chatbot has become fashionable. It seems like a modern term or idea, but nothing could be further from the truth. The 1960s, a time when many important events took place, like the Cold War, demonstrations against the Vietnam War or the assassination of John F. Kennedy, Malcolm X and Martin Luther King, marked the dawn of chatbots.
ELIZA was the first Chatbot. It was the year 1966 when its development was completed at the Massachusetts Institute of Technology (MIT). Its function was to maintain a coherent conversation with a user, but it was a very simple system. Its programming was based on the search for keywords and returned predefined phrases that matched the keywords found. When the system did not find any keywords, and was therefore unable to give an answer, it reformulated the expression received as a question or reflection. The conversation quickly became meaningless.
The chatbot PARRY was developed 6 years later in the Californian university of Stanford. This bot simulated a patient with paranoid schizophrenia. It was described as “a version of ELIZA with attitude”. Variations of the Turing Test, carried out by a team of psychiatrists, were applied to this system. The psychiatrists were successful in identifying whether the interlocutor was the bot or a human only 48% of the time. PARRY got to communicate with ELIZA on several occasions. One of these “conversations” took place through ARPANET, the US military network that was precursor of the Internet.
1988, 16 years since the birth of PARRY, Jabberwacky appears. It is a more sophisticated project, which purpose is to pass without problem the Turing Test. But it goes even further than its predecessors. Now the interaction does not take place through the keyboard, but through voice commands. Its creators had the idea of incorporating it into objects, such as a robot, something that is very familiar to us today.
1992, the year of the Olympic Games of Barcelona, Creative Labs, a brand then well known for its sound blasters, creates Dr. Sbaitso, an acronym for Sound Blaster Artificial Intelligent Text to Speech Operator. The software takes the role of a psychologist, using the sound technology of the time to synthesize his answers and emulate, albeit poorly, a conversation.
PARRY SIMULATED A PATIENT WITH PARANOID SCHIZOPHRENIA
ALICE (Artificial Linguistic Internet Computer Entity) came out online in 1995, on an Internet still in its infancy. Using the JAVA programming language and heuristic patterns, it made it difficult for the user to know if the conversation was with a computer or with another person.
We leave behind the 20th century and go on to the year 2001. SmarterChild operates on the most important messaging platforms of the time: MSN Messenger, AOL Instant Messenger and ICQ. Not only was it able to have a conversation but he could execute a series of commands. It is one of the forerunners of the enterprise bots we know today. It already included advertising in its conversations and even had a paid version, which was not very successful among users, and led to the temporary closure of the company responsible for it, Colloquis Inc., which later reopened, this time completely free.
In 2006 the multinational IBM launches Watson, named after the founder and first president of the company, Thomas J. Watson. This tool, responsibility of the DeepQA research team, is capable of having natural conversations using external sources of information such as encyclopedias, dictionaries, thesauri, news articles and literary works, among others. It also thrived on the knowledge acquired from Machine Learning. In 2011 Watson participated in an American television contest called Jeopardy! He competed against two humans and won the first prize of one million dollars. This amount was donated by IBM to two NGOs.
2010, the turning point. Apple presents Siri, the evolution and forerunner of all current chatbots. As usual Apple made significant improvement by redefining what the user expects from an automatic system. Such has been the impact that Siri has had in society that it has appeared in several series and movies. For example, in a chapter of the popular series The Big Band Theory in which one of its main characters, Raj, was very enthusiastic about Cupertino’s assistant and got to know her in person in a dream. This development is undoubtedly a turning point for virtual assistants.
Siri’s integration has been exponential. Although it was initially limited, today Apple gives permission to app developers to use Siri’s potential, allowing the creation of tools that respond to voice in a completely natural way. In addition, it allows the interaction with the home automation system. The user can ask Siri to turn on the lights or tell it which TV series he wants to watch. Siri is present in almost all Apple devices: iMac, MacBook, iPhone, iPad, Apple Watch, Apple TV and in its recent intelligent speaker, the HomePod.
And then came the others. In 2012, to take the lead, Google presents its version of Siri, Google Now. Like Apple, the company of the famous search engine will begin to include its virtual assistant in its devices and in Android. The latest addition to Google Now has been its recent range of speakers, Google Home.
2015 is the year of Microsoft and Amazon. The first launches Cortana, using the potential of its search engine Bing, while the Seattle company includes in its Amazon Echo speakers Alexa’s technology, which allows users to interact with their device through natural voice conversations.
Each assistant has a different voice command to start “listening”, so Siri listens to us when we say “Hey Siri” and Google Now responds when we say “Ok Google”. Alexa listens upon hearing its name, which is not practical if someone in the house has the same name, because every time it hears this word it activates. Amazon had to explain how to modify this default option to avoid continuous nuisances in these cases.
A year later, Facebook allowed its users to create bots on its platform. This opened a world of possibilities in terms of chat between companies and users on the social network. By June of that year, 11,000 bots had been created. Other platforms, such as Twitter, will later have their own bots, with their own tools or those of third parties.