The origin of chatbots can be found in Alan Turing’s groundbreaking book Computing Machinery and Intelligence, which explores the question of whether machines have the capacity for thought.
These traits would be reflected in the so-called Turing test. In its most basic form, the test is given as an imitation game. In the trial, a human interrogator interacts with various systems and individuals via an interface. If the interrogator is unable to distinguish between machines and people, the Turing Test is deemed to have been passed.
As one of the first applications of natural language, Joseph Weizenbaum, a computer scientist at the MIT Computer Science and Artificial Intelligence Laboratory, created the chatbot ELIZA between 1964 and 1966. ELIZA’s initial goal was to imitate the conversational style of a Rogerian psychotherapist using straightforward pattern matching and a response built on a template (prewritten scripts). The development of a natural language bot that could pass the Turing Test gained attention after ELIZA. The Loebner Prize is an illustration of a competition expressly created to implement the Turing Test and builds on both the expanding interest and developments in AI. It was established in 1990 by Hugh Loebner and the Cambridge Center for Behavioural Studies in Massachusetts (U.S.).
In the late 1990s, American scientist Richard Wallace expanded the pattern-matching methods used in ELIZA and improved NLP capabilities to build ALICE (Artificial Linguistic Internet Computer Entity). ALICE, also referred to as Alicebot, is notable for its creative application of Artificial Intelligence Markup Language (AIML), an XML framework for creating software agents that can understand natural language. Wallace published the first version of AIML in July 2001 and the Pandora API, on which ALICE is based. ALICE was the recipient of three Loebner Awards, in 2000, 2001, and 2004.
In parallel with ALICE initiatives, British programmer Rollo Carpenter was developing “Jabberwacky.” To mimic “regular human speech in an entertaining, humorous, and natural way,” Jabberwacky was created. Based on a dynamic database of thousands of online human interactions made possible by the development of the Internet, Jabberwacky was able to process responses. Under the guises of George and Joan, Jabberwacky won the Loebner Prize in 2005 and 2006, respectively. A new variation of Jabberwacky called “Cleverbot” was introduced in 2008. Similar to Jabberwacky, Cleverbot is made to learn from human conversations (more than 150 million to date, according to Wikipedia). It bases future questions and answers on previous interactions.
Conclusion
To create an AI that can pass the Turing Test is the project’s goal. It is intended to communicate with users by simulating human contact. It is not meant to be used for anything else. In contrast to normal AI programmes, the learning technique is intended to be utilised for entertainment rather than for computer support systems or commercial representation. Recent developments have made it possible to layer a more scripted, controlled approach on top of the general conversational AI with the goal of fusing the best elements of both approaches. Use is now taking place in the sales and marketing industries.
Additionally, the programme will eventually switch from a text-based system to a voice-operated system that only receives auditory and other sensory inputs in order to learn. Last but not least, according to its creator, we can include it into home things like talking dogs or robots to be useful and fun while keeping people company.
