I haven’t played computer games in a long time, not since the days of Quake 2, but I’ve kept an eye on developments in the field and I think I can take an educated guess on where the state of the art will be relatively soon.
So imagine you are playing a first person action-adventure game that takes place around New York City. Your friend gets killed in building A, you investigate, find clues that lead you to building B to meet certain people, etc. There’s a fairly linear storyline that you can follow and it will lead you to a conclusion that wraps things up.
But lets say you don’t feel like following the main plot. Nothing special about that, lots of games give you freedom to wander around and explore. What I’m taking about is taking this to the next level not only in scope, but in detail and interactivity.
So you are in virtual NYC. What would you be able to do? What about taking a cab to New Jersey, going to Newark to buy a plane ticket, fly to Paris and have a drink under the Eiffel Tower, then go get another plane ticket, fly to Saudi Arabia and visit Mecca. Or go to Tokyo and find a karaoke…
Why not? With a game engine sophisticated enough, all of these things could be generated almost automatically as long as you have lots of raw data. During game development, you would feed it tons of detailed maps, satellite photos, encyclopedia information, government statistics, architectural blueprints and demographic information. For the details, you could probably feed it a few terabytes of public domain geo-tagged photos and videos (from sites like Flickr and Youtube, or whatever we have a few years from now) that would help with the appearance of buildings in various places, how people dress in different parts of the world, local plants and animals, parks and waterways, hills, mountains, etc. The blind spots could be extrapolated or filled in by the programmers.
Non-playing characters (NPCs) could be sprinkled around realistically based on demographic information, and those close to the player could act intelligently based on advanced AI rules and those out of sight could still evolve in statistically probable ways, so that the world isn’t completely fixed and if you play the game long enough, things will change. NPCs with whom the player interacts often with could get their own storylines and become more “human” while others would just melt back into the background statistics once out of sight.
I think this would sell. People have always been attracted by exploration and travel. Even in game with severely limited worlds, you can see players exploring all corners, trying to go farther even if they know that the next goal in the game probably isn’t in that direction.
I don’t think all games will be like this; it’s not because you can do something that you should. In fact, the first piece of software to be like this will probably be designed specifically to use that vast world. Something like a Massive Multiplayer Online Role Playing Game (MMORPG), because that’s the best way to justify the cost of building it. But later on, I’m pretty sure that other games will go that far just because they can do it relatively easily and it will be a good PR move (“our world is vaster, more detailed and more interactive than yours! Total freedom, for real this time!”).
After all, everything I’ve described above is just an evolution of techniques that are already used on a (much) smaller scale. It’s actually fairly boring compared to what we will be able to do after a few revolutionary things are developed (connecting directly to the nervous system, for example). You could get real virtual freedom sooner than you think.
Thanks to Ben Boyd for reading a draft of this and giving me insightful feedback.
Update: Here’s one upcoming technology that will make games much more immersive (and fun!). Please watch until the demo starts in the last 1/3: