Kent Stevens

When people play computer games is the computer choosing responses or is it just responding? If you play the computer at Chess, Poker, or Tic Tac Toe is the computer making decisions or is it just having certain responses being triggered? If you play the computer in a car race or a beat 'em up scenario, is your opponent choosing?

You could have the computer being said to be doing things decisonally. So you could say that the computer decided to play a certain Chess move. It chose this move after looking at alternatives and evaluating the consequences of the various moves. It has a weaker form of "rationality", in that it's moves are informed and they are not crazy, if the goal of the computer is to win.

You could also have the computer doing things without reference to making decisions. The computer could be said to be doing things physically. The computer played a certain move. It considered alternatives, processed information and came up with one move being the best move. The computer had certain rules being triggered. In this context the computer generates, makes, or produces moves.

I believe that the decisional and the physical interpretation both have a degree of validity to them. However, I would tend to prefer saying that the computer simply does things, as opposed to saying that a computer decides to do certain things. I am slightly uncomfortable with expressions that involve the computer choosing. But I do not mind that much if the alternative decisional interpretation is also used.

I think that people are analogous to computer games in that you can describe what they do decisionally or physically. In terms of what people do, I prefer the physical interpretation, but I also use the decisional viewpoint.

ax

The computer has a list of responses, and depending on what you you, you activate one-hence it seems like the computer "choosing" a response.
AI isn't that developed yet! (thank G-d!)

K

I would say that the computer chooses. It's just not self-aware. It doesn't observe the choice it makes like we do when we make decisions.

wade-w

The computer is only running software programs that are designed by a human. Any "choices" made in a game are programmed responses, made in accordance with some logic that was dictated by a human when the software was designed.

acronos

The easy answer is "no" because choice could be defined to require a human or conscious being. I think your question is far more enlightening than such an answer. My answer is "yes" for the sake of consistency. I believe a human making a choice is massively more complicated than what a river does when it chooses to carry out more limestone than granite. However, at its core, it is the same thing.

The human centric aspect of choice is an illusion. Choice is a physical action whether it is made by a human or a computer. I believe that after we discover what the correct physical mechanism is for choice, it will be much harder to view choice the way it was viewed before. We will find that choice is a direct result of external stimulus and internal states and memories. For any given stimulus and neuron configuration it will be possible to predict the outcome. As such a human choice is only a vastly more complicated version of what the computer is doing or what a river is doing.

GodLessWarriorTM

Many of the more advanced games will learn how you play a specific game so that each time you play the game the computer will check into its recorded database to determine which decisions you are more likely to make. Simpler games, such as chess (not that simple) is based on algorithms that analyze the current position of all of the pieces on a given board and then deciding you best posible move. Once it calculates what your best move could possibly be it attempts to calculate how it will counter. It keeps performing a best/worst case scenario for about the next 5 moves. The more advanced algorithms will go even further than that.

The more I think about it though I have a harder time determining why a computers decisions are any different then ours. As humans we make decisions on past experiences so as the predicted outcome is as much to our advantage as possible. A computer can also make decisions based on past experiences and will make decisions so that the outcome will be in its advantage.

Computers can now be programmed to have neural network simulations to create more advance AI than ever before. The first game off the top of my head that I know of that utilizes this technique is QuakeIII, the popular multiplayer first person shooter found being played in offices all over the world (except in Greece of course. Whats with those people???) The neural networks retain information by association. For example lets assume we have a small network of neurons that represent "red". Everytime a red object is detected that object, lets use a "car", reinforces the "red" network by making more connections to it. In the future if another "car" is detected the car network will become stronger. This is a very basic example of how we remember things. If the networks are not reinforced by stimuli from the outside world they begin to deteriorate, which is why the older we become the more we forget. Now, just imagine millions and millions of these neurons connecting to form a single massive network. A network so complex that it makes us look like we are "choosing" things on our own. But like Acronos was saying, the decisions are based on how your network is setup. So how is that any different than a computer?

[ September 07, 2002: Message edited by: GodLessWarriorTM ]</p>

1sec

Choose = select through decision or preference.

So, no.

K

Why is it not a selection?

K

I'm sorry, I meant to say:

Why is the selection not a decision?

Vibr8gKiwi

It's never quite as simple as "the computer has a list of responses, and depending on what you do, it activates one." Computer reasoning these days is much advanced than that.

For anyone interested in how complex even a trivial application of computer decision making can go, check out the <a href="http://www.cs.ualberta.ca/~darse/rsbpc.html" target="_blank">RoShamBo Programming Competition.</a> (that link might not be useful to non-programmers, however the link below to an explanation of one of the programs should be fairly accessible to everyone I think).

It's a programming contest where two programs play games of "rock, paper, scissors" against each other. Even for such a trivial game, some of the programs use incredibly intricate decision making. Even for me as a programmer, trying to follow the logic some of these programs use can be a mind-straining experience.

For example, one of the most interesting (and strongest) programs is called "Iocaine Powder" (after a scene from "The Princess Bride"). The program tries to reason like the character Vizzini in the scene where two opponents play a game of wits to the death: "Now, a clever man would put the poison into his own goblet, because he would know that only a great fool would reach for what he was given. I am not a great fool, so I can clearly not choose the wine in front of you. But you must have known I was not a great fool, you would have counted on it, so I can clearly not choose the wine in front of me! ..."

The forms of double and triple guessing the program uses are explained at
<a href="http://www.ofb.net/~egnor/iocaine.html" target="_blank">Iocaine Powder Explained.</a> After reading that I'd hope there'd be more appreciation for how much reasoning can go into even the most trivial computer decision making.

[ September 07, 2002: Message edited by: Vibr8gKiwi ]</p>