Getting Started with Artificial Intelligence

Chapter 3. Chatbots

Dr. Ashok Goel teaches a class called Knowledge-Based Artificial Intelligence every semester at Georgia Tech. It’s a massive class of around 300 registered students. Understandably, being a teaching assistant (TA) for this course can be quite the challenge as 300 students post roughly 10,000 messages in the online forums. But during the spring semester of 2017, one particular TA was especially good at her job. She answered the students’ questions with excellent efficiency. She would even make herself available during late hours the night before a deadline. Naturally, the students loved her and gave her rave reviews.

You can probably guess how the story ends. At the end of the semester, most students were surprised to find out that the TA, named Jill Watson, was actually a chatbot. Dr. Goel and his team built Jill by tracking down all the questions that had ever been asked on the course’s online forum (about 40,000 postings in all). They then trained Jill to answer these questions using IBM Watson. Jill wasn’t very good at first, but she learned from her mistakes and was eventually giving answers with 97% certainty.

It’s stories like this that sparked the interest of developers and entrepreneurs and made 2016 the year of chatbots. In this chapter, we will explore a few aspects of this topic. We’ll first talk about what a chatbot is and why now is a good time for you to build one. We’ll then share a few important considerations for creating a successful chatbot, ...