The computer asks the user a series of yes/no questions, then gives a guess. If the guess is incorrect, the computer asks the user for the correct answer and a question to differentiate between the correct answer and the wrong answer. Over time, the computer will become better and better at guessing.