Artificial Intelligence: What Computers Can Do

The field of artificial intelligence has long been one of the most fascinating areas of computing to the general public. Ever since we have had digital computers, there have been predictions that "within 10-20 years" computers will be able to perform problem solving tasks as well as or better than people, and these predictions continue today. Artificial intelligence developments have been in the news consistently in recent years (IBM's Watson defeating the best humans in Jeopardy, Google's AlphaGo defeating a highly ranked player in the strategy game Go, and a variety of companies making progress on developing self-driving cars). Due to these developments and others, a number of leading computer scientists and tech company executives have suggested that we are now at the point where we need to consider how to place limits on the autonomy of artificial intelligence to ensure we avoid an apocalyptic future where we lose control of the machines and they become a threat to humanity.

In this seminar, we study many of the key techniques of artificial intelligence to gain an understanding and appreciation of the field. The goal is to understand what is currently possible and to be able to use AI tools to solve problems. The following are among the techniques and topics that will be studied: artificial neural networks, so-called deep learning, computer game AI, the Turing test, and model construction. By the end of the seminar, students will have an understanding of how the algorithms work, the current limits of artificial intelligence, and will develop their own opinion of what AI can achieve "within 10-20 years".

Because computer programming is a key background skill needed to understand the techniques, the seminar also develops basic programming skills. Each day of the seminar will be a mix of looking at a key idea in artificial intelligence, learning about computer programming, and making an AI project. Students will complete an AI project throughout the course of the week. Students can choose to improve on one of the in-class projects or to do a more thorough background study of an area of AI that is of interest to them. The seminar does not require previous programming experience. All that is required is that the student is interested in learning more about AI.

Dr. Jeff Kinne, Associate Professor
Department of Mathematics and Computer Science
Indiana State University