Educators’ concern and interest in artificial intelligence led TechnoKids to create TechnoChatbot AI. This is a technology project in which students become chatbot developers. They use different programming apps to create a variety of unique and useful virtual assistants. An extension activity in the project explores the history of chatbots. The purpose of the assignment is to give students a background in artificial intelligence and chatbots.
What Are Chatbots?
Chatbots can imitate human conversations. They use technology, called natural processing language (NLP) to understand text and spoken words. They can be programmed to respond like a person. For example, they can be trained to ask suitable questions, give responses, and express emotion. A timeline of the history of chatbots reveals how long people have studied artificial intelligence and it development to the present day.
History of Chatbots
1950 Alan Turing and the Turing Test
Over 80 years ago, Alan Turing wrote an article in the journal MIND. In it, he proposed the question, “Can machines think?”
To study the problem, he created an imitation game. In it a judge asked questions to both a human and machine. The machine tried to answer like a human to see if it could fool the judge. It did not. This became known as the Turing Test.
Today, despite the many advances in artificial intelligence, most computers cannot pass the Turing test. People can usually tell they are not human. But…they are getting better at imitating us.
1966 – Joseph Weizenbaum and ELIZA
Joseph Weizenbaum created a program called ELIZA. He wanted to explore if a computer could have a conversation like a real person. ELIZA acted as a therapist. It asked open-ended questions to a person that would then type answers.
ELIZA did not understand the person’s problems. Instead, it used if-then logic, keywords, and stock phrases to express empathy. Many users liked chatting with ELIZA, even though it was a machine.
ELIZA was an early natural language processing program. Today, machines can understand written and spoken language. However, the emotions they express are not human.
Have a look at this video to learn why ELIZA is significant in the history of chatbots.
1995 – Richard Wallace and ALICE
Richard Wallace created ALICE or Artificial Linguistics Internet Computer Entity. It was one of the first chatbots that could realistically follow the pattern of human conversations. It was programmed with thousands of conversation rules and topics.
To make ALICE chat, Wallace, with a community of developers, created the free software AIML or Artificial Intelligence Markup Language. AIML included a knowledge base of categories with questions and answers. It made ALICE seem smart and friendly.
Since AIML is opensource, many chatbots today have been created thanks to ALICE. In fact, AIML is used to create realistic avatars that can play games, teach, or help customers.
Kuki is a good example of a chatbot that was made using AIML. Try chatting with it. Here are some suggestions.
- What do you know about tigers?
- Do you like pizza?
- Are you real?
- Tell me a joke.
2010 – Apple and SIRI
Apple created Siri to be a voice-controlled personal assistant on their devices, such as the iPhone or iPad. Siri became so popular it is still used today. This is because it can instantly answer questions, play music, provide navigation, phone a friend, and more!
Siri can recognize speech and talk like a person. Since it has access to the Internet it seems to know everything. Plus, it can use built-in apps such as mail, contacts, and messages to do a task. This makes it very helpful.
Today there are many virtual assistants such as Apple Siri, Amazon Alexa, and Google Assistant.
2022 – OpenAI and ChatGPT
OpenAI developed ChatGPT. It is an artificial intelligence chatbot that mimics human conversations. It can customize the length and style of its response. Not only that, but it can do many writing tasks. For example, it can create poetry, compose songs, and code a program.
Most AI Chatbots such as ChatGPT are available online. Anyone needs to register for an account. Typically, teenagers require their parents’ permission to use the platform. Once signed in, you type a message. The chatbot then replies. You can rate the response, which will help it improve.
Try it out here! Here’s a sample output from ChatGPT.
|Write me 50 words about polar bears in the style of a sports announcer.|
|Ladies and gentlemen, get ready for a chilling spectacle as the Arctic’s finest athletes take center ice! The mighty polar bears, nature’s powerhouse, glide effortlessly on the frozen terrain, showcasing their brute strength and agility. It’s a gripping display of survival and grace in the ultimate icy showdown.|