Some key areas of technology that will change over next few years of the industry:
– adaptive content: can’t scale the art teams as rapidly as would be needed, plus memory bandwidth/capacity on consoles is even more critical in forthcoming generation => therefore need content that is more detailed but less specialized (i.e. can be redeployed in different scenarios in the game). This subsumes procedural generation of textures, models, cities, characters and also includes procedural animation and AI for those models/textures.
– real time content tools: boy is this a black hole in the industry right now. You can create a crate in max/maya/xsi but what tool do you use to attach a script to it so that it blows up and emits a sound when collided with by a grenade? The industry is partially crippled due to the deadlock between coders and artists – this dependence needs to be broken to facilitate efficient team operations.
– Speech Synthesis & Recognition – cited by many as one of the next big hurdles to immersion and emotional content delivery. Adaptive dialogue generation to get rid of the Monkey Island “choose 1 of 4 reponses” approach.
– AI – the great challenge. The term encompasses too much really – but somehow developing workable solutions that allow combining of low level but plausible steering & interaction behaviors for characters with high level story arc / narrative direction scripts… well that’s just hard, but this is the crux of interactive story telling. Fool the gamer into thinking they have choices, but in fact bring them along a realtively linear story arc. Your allies in this are the environment and the characters that you interact with. Make it persistent and consistent and plausible and you’re on track.
These are all very tech focussed I guess.