Summary: CS680: Assignment SB­1 Q1
Questions:
This preparation assignment has a preponderance of ``essay'' or ``thought'' questions. Do not be
intimidated by these. You should respond to them with a few pages of text overall. The main idea is
for you to think about the reading and the topics we will discuss in class. It doesn't matter how smart
you are---to say something interesting you need to think about it in advance. And writing down your
thoughts, even quickly, is the best way to make your thoughts more concrete. Thus, the point of these
preparation assignments is to get you to think about the topic in a relatively concrete fashion. There
are no right or wrong answers, and your writing will not be critically evaluated.
Question 1.1 Characteristics of an Intelligent Agent. What are the essential features of an intelligent
agent? What activities and abilities must all such systems share?
Question 1.2 High­ and Low­level AI. AI has usually emphasized high­level intelligence, such as a
doctor diagnosing a patient or a master chess player deciding on a move. In this course, however, we will
emphasize lower­level intelligence, such as may be involved in more mundane things like walking across
a room or making breakfast. How are these two kinds of intelligence different? How are they the same?
Which direction of approach to AI seems more appropriate to you? Why? What other thoughts do you
have in response to Chapter 1?
The following questions pertain to the Tic­Tac­Toe problem and the reinforcement learning (RL)
algorithm for playing it described in the draft chapter 1.
Question 1.3 Self Play. Suppose, instead of playing against a random oppponent, the RL algorithm

Source: Anderson, Charles W. - Department of Computer Science, Colorado State University

Collections: Computer Technologies and Information Sciences