Tetris and the nature of online gaming
So, I just happened across TetrisFriends.com. I've been climbing the ranks in Sprint 5, Battle 2P and Battle 6P. It's been fun having human opponents to play Tetris against.
But when I lost my first Sprint game, it occurred to me that the experience I had might have been easily generated by an AI. There's very little interaction in the game, as there's no chat and what really matters is how fast you go.
All an AI would have to do would be to slowly increase its speed until I lose, then adjust it up and down to give me good challenges and some wins to keep me coming back. I know it's not all that simple. The people at the site can list their names and join friends lists and such.
But given the simplicity of Tetris (which is one of its main strengths), there really isn't much room for individual strategies. Seeing different tactics used against you is one of the cool things about playing other games against people.
Also, TetrisFriends seems to be rather successful, so there's a large pool of people to play against, which makes sure the difficulty curve is very smooth. While this is in principle a good thing, it makes your opponents rather faceless at the same time.
But then it occurred to me: aren't all online games like this? Except for chatting, which AIs can't replicate yet, couldn't most online game experiences be replicated with AIs?
Well, not really. It depends on the complexity of the game. When you play something like Quake 3, and there's a bot in there, it's really obvious. But Q3 is a 10-year-old game; perhaps with better AI it could work. But then there's the issue of teamwork. Getting in a squad in Killzone 2 and working together with the others yields some very interesting teamwork dynamics. This sort of thing is still beyond the reach of current AI techniques, at least with the limited hardware available to home console games.
But in the future, perhaps AI will advance far enough to give a really good approximation of human behavior in games. What then? There is also the wider question of what happens when AI can replace an even greater range of human behavior? I can imagine people needing to interact with each other less and less, and humanity disappearing through disconnection into a new world of computer sentience. Or maybe the same technology will just better enable people to share their imaginations with each other. We can only speculate at this point.
My apologies for this post turning into a ramble. But that just proves this was written by a person.
Or does it?
But when I lost my first Sprint game, it occurred to me that the experience I had might have been easily generated by an AI. There's very little interaction in the game, as there's no chat and what really matters is how fast you go.
All an AI would have to do would be to slowly increase its speed until I lose, then adjust it up and down to give me good challenges and some wins to keep me coming back. I know it's not all that simple. The people at the site can list their names and join friends lists and such.
But given the simplicity of Tetris (which is one of its main strengths), there really isn't much room for individual strategies. Seeing different tactics used against you is one of the cool things about playing other games against people.
Also, TetrisFriends seems to be rather successful, so there's a large pool of people to play against, which makes sure the difficulty curve is very smooth. While this is in principle a good thing, it makes your opponents rather faceless at the same time.
But then it occurred to me: aren't all online games like this? Except for chatting, which AIs can't replicate yet, couldn't most online game experiences be replicated with AIs?
Well, not really. It depends on the complexity of the game. When you play something like Quake 3, and there's a bot in there, it's really obvious. But Q3 is a 10-year-old game; perhaps with better AI it could work. But then there's the issue of teamwork. Getting in a squad in Killzone 2 and working together with the others yields some very interesting teamwork dynamics. This sort of thing is still beyond the reach of current AI techniques, at least with the limited hardware available to home console games.
But in the future, perhaps AI will advance far enough to give a really good approximation of human behavior in games. What then? There is also the wider question of what happens when AI can replace an even greater range of human behavior? I can imagine people needing to interact with each other less and less, and humanity disappearing through disconnection into a new world of computer sentience. Or maybe the same technology will just better enable people to share their imaginations with each other. We can only speculate at this point.
My apologies for this post turning into a ramble. But that just proves this was written by a person.
Or does it?