(Interactive chatbots) Thanks to the development of artificial intelligence, natural language processing technologies, and data science, chatbots are becoming faster and more interactive. In the 1960s, technology advanced. big.
The latest development in the field of interactive chatbots today is ChatGPT, developed by OpenAI, which is a dialogue-based chatbot model capable of understanding human speech, with the ability to write and debug code, admit errors and reject inappropriate requests. answer.
Last Monday Google announced the first beta version of Bard - Google's answer to the ChatGPT bot - based on an advanced language model developed by Google (LaMDA), and Microsoft announced that updated versions of the Bing and Edge ChatGPT search engines will be integrated into your browser.
All this leads us to wonder where this crazy race between major technology companies is going, and how have interactive chatbots evolved over the past 60 years into what we see today?
1- Eliza in 1966:
The origin of interactive chatbots lies in the scientific research paper titled (Computing Machinery and Intelligence) published by the British scientist (Alan Turing) in 1950, after which he developed a test later known as (Turing test) in the following decades. , the test had a significant impact on the debate about artificial intelligence.
Alan Turing focused his research and articles on the fact that machines can also think and have their own intelligence. He is a pioneer in AI standards, and once said, "If a machine can imitate a human and its behavior causes other people to engage in a real-time conversation that they think is interacting with a human rather than a machine, then the machine has intelligence."
But the first interactive chatbot as we know it today appeared in 1966 by Joseph Weisenbaum, a professor and researcher at the MIT Artificial Intelligence Laboratory, based on research by Alan Turing.
Joseph's goal is to enable dialogue between a computer and a human user by identifying keywords or phrases and asking follow-up questions of those keywords.
Elisa, a simple program of just 200 lines of code displayed on a Tandy machine and then developed by Radio Shack, responded to messages based on recognized keywords that were sent to it.
Eliza is the first chatbot in computer history, even if the term chatbot was not coined at the time. The bot works on the basis of pattern matching and replacement, making it easy for it to select keywords or phrases from an input in order to reproduce answers with those keywords from the pre-programmed answers. It made the early adopters feel like they were talking to someone who understood their point of view.
Example: When someone says, “Mom makes delicious food. The robot types in the word ‘mama’ and responds with an open-ended question: ‘Tell me more about your family.’ Although the process is automated, it creates the illusion of understanding and interacting with real people.
Weisenbaum explained that his main goal was to show the superficiality of human-computer interaction, but he was surprised to find that many users were so immersed in interacting with Elisa that they completely forgot that it was a program without real intelligence. Artificial intelligence - devoid of any feelings or emotions.
You could try to communicate with the Elisa version with some JavaScript improvements; Follow this link. But you have to realize that you are talking about software that was developed in the 60's.
2- Robot Barry in 1972:
Stanford University psychologist Kenneth Colby developed the Barry robot in 1972 to simulate a conversation with a paranoid person. Unlike ELIZA, PARRY has a more advanced speaking style and often responds accurately to written communications.
at the 1973 International Computer Conference; Vint Cerf - a computer scientist and one of the pioneers in the field of the Internet - decided to have a conversation between (Barry) and (Eliza) which, despite its simplicity and truth, was at the time the first conversation between two artificial intelligences in history.
3- Dr. Center The 1992 SBAITSO robot:
In the 1990s, chatbots were capable of talking to humans, such as B. chatbot (DR. SBAITSO ) DR in 1992. SBAITSO was designed to talk to humans and is known for its eerie digital voice that, while similar to previous chatbots, is nothing like Human.
In 1992, sound card company Creative Labs released Chatbot (DR. SBAITSO)—short for "Sound Blaster Acting Intelligent Text to Speech Operator"—a program designed to display digital audio that home sound cards could produce in a completely unrealistic quality.
The program talks to the user like a psychiatrist, although most of the answers are "Why do you feel this way?" Unlike any kind of complex interaction, his usual response when confronted with terms he doesn't understand is, "That's not my problem."
There is also a Prody Parrot-based version of this program for Microsoft Windows, which has a more detailed graphical user interface.
4- The A.L.I.C.E robot in 1995:
1995; Richard Wallace - computer scientist and inventor of the Artificial Intelligence Programming Language (AIML) - develops A.L.I.C.E. , a natural language processing conversational bot inspired by Joseph Weisenbaum's classic ELIZA bot.
The bot approach (pattern matching) is a major upgrade over previous chatbots. Updated A.L.I.C.E. It takes hours. It earned him three Loebner Prizes in 2000, 2001 and 2004 despite him failing the real Turing test.
5- Robot Smarterchild in 2001:
In 2001, ActiveBuddy developed an intelligent bot (Smarterchild) to integrate with customer service software. Then it spread widely through the global SMS network: AOL IM, MSN Messenger and Yahoo Messenger. Serve users similar to the current voice assistants Siri and Cortana.
It became very popular among users and attracted more than 30 million users on AIM, AOL, MSN, and Yahoo Messenger instant messengers during its lifetime.
ActiveBuddy changed its name to Colloquis and focused on developing chatbots for customer service, but Microsoft then bought it in 2007 (SmarterChild), which was seen as the starting point for voice assistants like Apple's Siri and Samsung's S Voice, far away.
6- IBM Watson 2006:
In 2006, IBM developed Watson to answer natural language questions, which was originally designed to answer questions from the popular question-answering software (Jeopardy).
In 2011, Watson faced two of Jeopardy's strongest players, Ken Jennings and Brad Rutter, and the program won first place and a $1 million prize.
Years later, IBM introduced Watson to businesses and hospitals, and in February 2013 the company announced that the first commercial application of Watson would be WellPoint, a lung cancer diagnostics facility at Memorial Sloan-Kettering Cancer Center. domain, great success.
7- Siri by Apple in 2011:
The voice assistant (Siri) Siri first appeared on the iPhone 4 in 2011, so interactive chat programs took another form, paving the way for voice assistants as we know them today, for example: Amazon's Google Assistant and Microsoft Cortana.
The main feature of the voice assistant (Siri) is that it is very good at handling the user's voice commands, such as: b.: call someone from the phone contacts or ask questions such as the weather, so it is very popular. Among users, and Apple is still around. A lot has been developed for its implementation and a lot of attention has been given to its launch.
Then, in 2014, Amazon introduced Alexa, its voice assistant, in the company's Amazon Echo smart speaker. Advertised as a personal assistant, Amazon Alexa can interact with users via voice with a high degree of fidelity and perform commands such as: play music, create to-do lists, provide weather information and traffic conditions, etc.
2016; Google introduced its voice assistant (Google Assistant) in the Allo messaging application, then Google announced in 2017 that its voice assistant would be available on all smartphones running Android 6.0 or higher, but at that time it was only available in English and German, and now it supports more 40 languages.
8- ChatGPT bots in 2022:
On November 30, 2022, OpenAI launched the first beta version of its interactive chatbot (ChatGPT), which is based on a large language model (GPT-3.5) for language prediction as it uses its own set of billions of words to process text and predict the next sequence of words in a given context to provide a human response. . In less than a week, ChatGPT beta reached 1 million users.
Chatbots (ChatGPT) ushered in a new era in artificial intelligence late last year, and have attracted widespread attention for their impact on the job market, as their ability to compose content, write code, and have very smooth conversations with humans has spawned proposals that can do. Soon we will be working in customer service, content writing, and even legal.
All this prompted the company (OpenAI) to provide a paid version of ChatGPT, which Microsoft hastened to integrate into its products, as it announced yesterday that an updated version of ChatGPT had been integrated into (Bing) Being Search Engine and Edge browser.
9- Google Call 2023:
After bots (ChatGPT) transformed the interactive chatbot market, Google entered the fray with the announcement of Bard, which is just a chatbot but will integrate its functionality into Google search results to provide succinct answers to the chatbot in addition to the usual offering of what users. Search for traditional search results. in Google Drive.
Before that, in August 2022, I submitted a beta version of an application called (AI Test Kitchen) based on a language model I developed (LaMDA). The company has begun allowing Android or iOS phone users with a Google account to download and sign up for the AI Test Kitchen app, which allows users to chat with LaMDA.
It is also expected that Google (LaMDA) will be integrated into many of its products in the future and launch new products based on its work.
finally:
Artificial intelligence still fascinates us today, its technology is developing very quickly and large companies are investing heavily in its development, it is a big step that seems to have borne fruit. What we've seen so far with companies rushing to integrate chatbots into their products suggests that we can expect more to come.