Spend a moment and watch the three squares move around the screen. Each square moves in a different pattern. One moves in a circular pattern, another in a square pattern and the third in a triangular pattern.
One of these squares is actively controlling the pattern it makes; it has the purpose of tracing out a particular pattern of movement. The other two squares have no purpose; they move as they do because their position is determined by an internal program. They are "output generators", not "input controllers". Output generators have no purpose; input controllers have the purpose of making a pattern that corresponds to an internal reference specification for that pattern.
Notice that it is impossible to tell, by looking at the behavior of the three squares, which square is moving purposefully and which are not. You can find out which is the purposeful square by applying a disturbance to all three squares using the mouse. The square with a purpose will resist these disturbances; the purposeless squares will not resist. This means that you can detect the square that is moving purposefully by looking for the square that is influenced least by mouse movements.
In this demonstration you are using The Test for the Controlled Variable to determine which of the three squares is controlling the pattern resulting from its movements. The square that is controlling its pattern of movements is behaving with the purpose of making a particular pattern of movement and resists disturbances to that pattern. The demonstration shows that you can't tell whether a behavior (such as a movement pattern) is purposeful without doing The Test to determine whether that behavior is under control.
Once you think you have detected the purposeful square you can press the mouse button (the cursor can be anywhere on the screen) and the actual purposeful square will be "filled in". Press the mouse button again and a new square (possibly the same square as before) will become the purposeful square and all squares will become unfilled. You can then try to find the new purposeful square.
Notice that that the movements of all three squares appear quite lively; there is nothing that distinguishes the appearance of purposeful from that of purposeless (mechanical) behavior. The only way to distinguish purposeful from purposeless behavior is by doing The Test to determine whether a variable (in this case, the pattern of movement around the screen) is under control
It may not be easy to tell which square is resisting your disturbances (mouse movements), especially if you use abrupt mouse movements to disturb the squares. Try using smooth, slow disturbances. Notice that two of the squares will be affected by your disturbances to some extent. Notice also that one of the squares seems to "do its own thing". This is the square that is protecting the pattern made by its movements from your disturbance.