I’d suggest game AI is not in a rut, per se – instead it’s standing on a slope, part way up a massive mountain, looking up at the distant peak, gone as far as it can go on foot, and now has to break out the mountain climbing gear and prepare for some very strenous, painful rock climbing. But imagine the views up there!

Also, a comment on Robin’s point about few women at AIIDE — that’s very true, but let’s not forget two of the primary voices that have had such an influence on framing, guiding and motivating this work for many researchers, particularly for interactive drama – Brenda Laurel and Janet Murray. And Marie-Laure Ryan is a newer voice with lots of new excellent analysis as well.

(1) I think it depends on how narrow your world is. If your world in any way includes people, it may be hard to get it just right, I think, because people do all sorts of rather complex things that look weird if almost-but-not-quite right. People-behaviors also seem pretty intertwined with each other, in the sense that if your characters are realistic humans in a few ways, it’ll be weird if other things don’t exist in your world at all. Though perhaps you can coax some apart. A sort of tangential point is that I think there’s a lot of interesting art in the “uncanny valley”. Whether this art takes the form of games people want to play, I don’t know, but I sure like the prospect of being able to unnerve people.

(2) Frankly, I didn’t think Bing’s talk was that interesting, but that may just be me. I could summarize most of it as “goddamn, Madden NFL sure does sell a lot of copies!” It’s a useful reality check, in that Madden NFL 2005 sells such a sheer number of copies that it explains why companies like EA aren’t that interested in taking a chance on more innovative stuff that, even if it “succeeded” by the standards of an innovative game, would probably still make less money than a month’s worth of Madden sales. The film industry has some similar problems I’d say, but in both industries, simply reminding yourself of the fact that a few blockbusters make a ton of money doesn’t take you much closer towards figuring out how to make something more interesting.

(1) About Will’s comment on the future behavioral uncanny valley… what’s the context here? It strikes me that this problem mainly takes form when unbounded… but within a space of defined expressive goals, such as one or three single behaviors, does the same problem persist? Granted, the game’s frame would have to do a damn good job divesting the player of interest in other forms of behavior. Wow, I hope at least some of that made sense.

(2) I’ve heard most of what Bing had to say before, save the seemingly important “AI matters.” What else did he say about this?

Incidentally, I used to work two doors down from this British Car Clinic Robin photographed. Tip: don’t buy a British car. This place was always overflowing.

As usual, I like Will Wright’s ideas a lot. “Drama is built on elaborate causal chains” – how very true. As long as “Why?” doesn’t work on everything, there’ll be no drama, IMO. Faking it – for instance, by treating “Why?” as “a provocation” – is not enough. Why? It limits character development tremendously. The result is likely to be a neurotic character. There’s no way to get depth like that.

observation #6: there are many applications of AI that are not opponents or human characters

“Right again, Robin.” SciFi, Horror, Fantasy are media genres chock full of non-human intelligent characters. Even cars can be Subjective Characters. Put that in a game and sell it.

I bet that the “meta AI” is going to be distributed across character AIs. Each character needs a director “built in” (a robot soul ;-); I don’t see a place for an external “manager/director” module. Rather, I see a hierarchy of characters. That, and a simple function at the fixed point. It is a matter of combining and nesting Storyforms.

Bing Gordon: “Gaming gets Pixared”. Yes, I think this will happen, but (I hope) on various scales. There’ll be smaller and bigger Pixars. Also, the Pixar content creation model might work in a distributed form – are “developer farms” really necessary? If they are, the Big Money guys are really at an advantage. Academics – could you please do some science on that?

But “engineers’ revenge”? Only if the engineers get really good at using Storyforms, or some other means of formalizing the procedural creation of meaning. Does he expect “engineer’s revenge” to happen at Pixar?

I wonder whether ScriptEase will be a solution for my authoring problems. It looks like it supports feeding a traditional “n++” content model. I’ll need a model, method, and tool(s) for creating “n^2″ type content on a large scale.

Ah, and Chris Crawford – “we fake thinking logically” – I seem to have observed this, too.

But: if we succeed in training young people to be both full-assed artists and full-assed programmers, are we in for a world of total double asses?