The Secret Lives of Chatbots

What is a chatbot?

Chatbots work just like AI robots. They are programmed to operate according to a predefined set of instructions. The more a chatbot interacts and learns new things, the smarter it becomes. They use the text-to-speech technology to deliver responses in a simulated human voice or text boxes.

Chatbots are typically used in dialogue systems for various practical purposes including customer service or information acquisition. Some chatterbots use complicated natural language processing systems. The system scans for keywords within the input and then creates a reply with the most matching keywords.

Today, most chatbots are found on virtual assistants such as Google Assistant and Amazon Alexa, or on messaging apps such as Facebook messenger or WeChat and other individual organisations' apps and websites.

Fun facts about chatbots

Chatbots are like artificial intelligence robots. They learn new responses out of interactions with the user in real-time, rather than being driven from a static database. Today, they combine evolutionary and real-time learning to optimize their ability to communicate based on each conversation held.

They also use a markup language called AIML which is specific to its function as a conversational agent and does not have reasoning techniques, but its functionality is purely based on pattern matching techniques.

The first robot test, the Turing test, was developed in 1950 by Alan Turing and was designed to determine whether AI is advanced enough to mimic human interaction completely. After his work, more developers saw the potential and growth of chatbots and later on in 1966, Joseph Weizenbaum created Eliza, the first active chatbot.

In 1972, PARRY, a computer program by Stanford University Scientist, Kenneth Colby modelled chatbots that imitated the behaviour of a paranoid schizophrenic. Later on, Jabberwacky, A.L.I.C.E & D.U.D.E were developed and are used in developing modern day chatbots.

Today, chatbots make their way into various industries. One of the earliest successful adoptions was by Nordstrom Rack who introduced a customer service / shopping assistant chatbot in 2016. Since, the helpful bot has served successfully over a million customers, guiding them through the vast store catalogue to the very item they had in mind.

Other popular fields for the chatbots are distant learning and healthcare.

Related article: What is a Chatbot and How to Use It for Your Business

Chatbots in education and e-learning

The gap between technology and education has been dramatically reduced by the use of innovative solutions such as chatbots. In the modern age, the implication of chatbots in a school environment creates an interactive learning experience for the student - similar to the one-to-one with the teacher.

From tracking the student's behaviour to testing and saving their progress, bots play a vital role in enhancing individual skills of the students. They also serve a significant purpose in encouraging and improving performance to students by sending regular reminders and notifications.

Today, chatbots are used in E-learning through:

  • Simulated guidance
    A simulated bot plays the character role of a teacher and guides the learner through the course and interacts with them through questions. Through this, a learner gets a more personalised education based on their current level of learning.

  • Quick summaries
    Typically, in a class environment, teachers use refresher quizzes and lessons to help students assimilate the information more effectively, but by using chatbots, teachers can now do the quick summaries by feeding the bot information at specific trigger points and intervals at the end of each milestone in a course.

  • Social learning
    Social learning is a significant cause and determinant of individual behaviours. Educational bots can be used to engage with students that are part of a larger group through social learning platforms (such as course applications) where students from different locations can interact with each other through pop-up discussions created through the bots.

Could chatbots make a difference in healthcare?

Today, almost everyone wants to ask the famous chatbot Siri about the weather or give instructions to Alexa to dim the lights, but would we be comfortable and trusting to talk to a medical chatbot?

How would you know whether or not to follow a bot's advice and how comfortable would you be following healthcare medications and instructions that came from an algorithm despite the initial set-up being approved by general practitioners?

The growth of chatbot industry is inevitable. While it promises to cost cuts, there is still fear that the chatbots, which are still at their AI infancy level, are risky and it is only safe to let the bots handle only non-vital tasks.

Currently, there are a few examples of AI-powered medical assistants:

  • Florence (the nurse) - this chatbot reminds you to take your medicine, monitors your health (and periods for women), can help you find health specialists and book medical appointments in your area.

  • Your MD  -  it acts like an assistant to a general practitioner, asks questions about symptoms and other aspects of your health which can help health professionals to identify a condition probabilistically. Then it sets up appointments, referring you to the most expert physicians for your problem.

  • Safedrugbot -  a chatbot messaging service that allows health professionals to keep record of possible side effects of drugs taken during breastfeeding and helps to keep breastfeeding women safe.

  • SimSensei  -  this bot uses voice and face recognition to mimic a therapist. It supplements human therapy and helps to interact with the patient at deeper levels.

To learn more about chatbots and their many applications, check out this infographic: Conquering The World – Chatbots Gone Wild.

This is a guest post by Josh Wardini. Josh is location incapable internet enthusiast, re-designer of the world around and bodybuilder trapped in a computer geeks body.