I have recently been reading about Massive, the technology used to create the climactic Battle of Helm's Deep segment in Lord of the Rings: The Two Towers, and I was struck by the similarity with game technology.
Massive, which stands for Multiple Agent Simulation System In Virtual Environment, was developed by Stephen Regelous for the Lord of the Rings movie trilogy. It is now available (more or less) as an independent product.
Rather than simulating crowds as simple particles - the state of the art until now - Massive uses a relatively sophisticated AI model to determine the actions of each individual crowd member, and then selects and blends motion-captured animations to display these actions. It also simulates both melee and ranged combat and movement on non-boring terrain, among other things.
The parallels with game technology are obvious: games have done, or tried to do, similar things for many years. However, that does not mean Massive is inferior to game technology. My impression is that it is superior to the AI in many games, although of course it does not have to perform in real-time, nor does it have to deal with irritations such as interactivity, fun, and 18 month development cycles.
One implication of this technology may be more interaction between animation software developers and entertainment software developers, in terms of people, know-how and actual technology. However, I don't think there is a general consensus on how to effectively use this kind of technology in games, although Take Two's State of Emergency and Electronic Arts' Lord of the Rings: The Two Towers give some indication of what can be done.
And yes, there are now three entertainment-related software development companies called Massive-something-or-other.