Skip to main content

Conference Report: AIIDE 2015

Earlier this month, I traveled to Santa Cruz, CA for two purposes: starting a new postdoc with the Expressive Intelligence Studio (which I'll be working on remotely for the time being) and attending AIIDE, a conference about AI in games, at which I was presenting a paper. I also attended two pre-conference events, the INT workshop on narrative tech, and the EXAG workshop on experimental AI in games.

In this post, I want to mention a few highlights among the many exciting things I learned at the union of these activities, and draw out some common themes.

1. The state of the art in computational narrative is social & emotional believability.

INT this year was joint with SBG, Social Believability in Games, and the themes of the talks blended so seamlessly that it was difficult to remember there were actually two workshops. Iolanda Leite gave the keynote on human-robot interaction; specifically, how to create conversational agents that would maintain children's interest over long-term durations, rather than the usual focus on a single interactive session. Her talk set the stage for other presentations on conversational NPCs, social mediation AI for MMOs and social networks, and the production of socially and culturally appropriate behavior.  I also had several chances to chat with members of the budding research group at American University consisting of UCSC alums Mike Treanor, Josh McCoy, and Anne Sullivan, who organized SBG, and learn about their work on the next generation of Ensemble, a scripting tool for social physics.

The underlying premise of all of this work seems to be that responsive narrative depends on socially believable characters. Meanwhile, social believability is still really hard! There are a lot of pieces to it, like cultural context, social norms within that context, the human interactors' personal/emotional context, the fact that humans remember, forget, and lie (which there was actually an EXAG paper about), and more.

So I'm of two minds about this premise. On the one hand, there are an awful lot of interesting problems in social believability, and I'm really excited to see what comes out of that area in the next few years (especially if epistemic modal logic might be involved). On the other hand, I think it's a mistake to suppose that advances in computational narrative can't be made without it. In some ways, stories are a lot simpler than characters, while it might be tempting to assume that good stories will naturally emerge from any set of sufficiently interesting characters (the premise of reality television), I find myself increasingly intrigued by the possibility of operationalizing approaches like the collaborative storytelling system Microscope, wherein the unit of content isn't a character interaction but an era, epoch, or historical event, each of which may involve characters as instruments, but more fundamentally, must fit together in a global way to meet the demands of narrative structure.

Somewhat tangential, but also related to social and narrative games: at AIIDE proper, Mitu Khandaker-Kokoris gave an amazing keynote (available to watch online) making the point that, often, what makes systems feel social is less about detailed digital realization and more about the gaps that we humans fill in with our willingness to suspend belief and project humanity onto said systems. Katherine Isbister also gave a keynote (unfortunately not recorder), centering on many similar themes, especially the idea that games can be vehicles for emotional and social bonding, providing context for her work in wearable game interfaces. I strongly recommend following Mitu and Katherine's groundbreaking work.

And speaking of social play and storytelling...

2. There's a lot of excitement right now at the intersection of human-driven collaborative storytelling and computer-driven play.

Ian Horswill gave a cool talk at EXAG on generating Fiasco playsets with Prolog, the upshot of which was basically "play Fiasco" (I still haven't), but I'm also enamored of the idea of using computational assistants for primarily human-driven improvy storygames.

Markus Eger spoke at INT about operationalizing PLOTTO, the "master book of all plots," and, at Ian's suggestion, a bunch of grad students and I got together to do a story jam where we semi-democratically made choices about which plot development options to accept until we reached a terminal node. Some notes and sample output here. An interesting thing we learned was that the plot developments don't necessarily come in causal order, but in many cases it is possible to read a development as further elaboration or backstory on a previously-described situation/event. This fact again touches on the idea of iterative and global, rather than linear and simulationist, story generation. (Of course, a downside of PLOTTO is that it's a 1-Markov chain in terms of expressiveness, i.e. it only maintains a history of the previous choice to influence future choices. So there wind up being a lot of loose story threads that never get called upon again later.)

I also spoke at length to various attendees about interactive/immersive/participatory theater. It seems like tons of games people know about, and are intrigued by, Sleep No More, the multi-room concurrent production in New York. (I still need to get my butt on an Amtrak to go see it, but I've heard and read plenty of descriptions at this point.) Pittsburgh's own thriving experimental theater scene has introduced me to Tamara, which I wrote about previously (on this blog and in my thesis), as well as Strata, a gauntlet of immersive/interactive vignettes, and Uncumber's suite of interactive work, including Her Things, a room full of explorable artifacts designed to be pieced together in pursuit of a narrative, held together by improvising actors. These productions serve as interesting points in a design space for collaborative storytelling, in which roleplaying, storygames, and computational fiction all share certain objectives -- blur the line between audience member and story-maker -- but which differ substantially in others (allow participants to alter the story fundamentally, or simply to choose a trajectory through it? allow participants to interact with one another? are participants characters or observers?). I'm very interested in how computational approaches can learn from recent participatory theater experiments and vice-versa.

---

While I have quite a bit more to say about the conference, I'll curtail the post here in the interest of length and thematic coherence.


Comments

Popular posts from this blog

Using Twine for Games Research (Part II)

This preliminary discussion introduced my thoughts on using Twine as a tool for creating prototypes for games research. I'll start with documenting my first case study: a hack-and-slash RPG-like setting where the player character has a record of attributes ("stats") that evolve through actions that turn certain resources (money, health, items) into others. I've selected this hack-and-slash example because it falls outside the canonical "branching story" domain thought to be Twine's primary use case, but it is not too much trickier to implement. It relies crucially on the management of state in ways that simple branching stories would not, but it does so in a fairly straightforward way.

If all goes well, this post may also serve as a tutorial on the "basics" of Twine (links + variables + expressions). In particular, I'll be using Twine 2/Harlowe, and I haven't seen many tutorials for this new version published yet.

To me, the main "…

Why I don't like the term "AI"

Content note: I replicate some ableist language in this post for the sake of calling it out as ableist.

In games research, some people take pains to distinguish artificial intelligence from computational intelligence (Wikipedia summary), with the primary issue being that AI cares more about replicating human behavior, while CI is "human-behavior-inspired" approaches to solving concrete problems. I don't strongly identify with one of these sub-areas more than the other; the extent to which I hold an opinion is mainly that I find the distinction a bit silly, given that the practical effects seem mainly to be that there are two conferences (CIG and AIIDE) that attract the same people, and a journal (TCIAIG - Transactions on Computational Intelligence and Artificial Intelligence in Games) that seems to resolve the problem by replacing instances of "AI" with "CI/AI."

I have a vague, un-citeable memory of hearing another argument from people who dislike the…

Using Twine for Games Research (Part III)

Where we last left off, I described Twine's basic capabilities and illustrated how to use them in Twine 2 by way of a tiny hack-and-slash RPG mechanic. You can play the result, and you should also be able to download that HTML file and use Twine 2's "import file" mechanism to load the editable source code/passage layout.

Notice that, in terms of game design, it's not much more sophisticated than a slot machine: the only interesting decision we've incorporated is for the player to determine when to stop pushing her luck with repeated adventures and go home with the current spoils.

What makes this type of RPG strategy more interesting to me is the sorts of decisions that can have longer-term effects, the ones where you spend an accumulation of resources on one of several things that might have a substantial payoff down the road. In a more character-based setting, this could be something like increasing skill levels or adding personality traits.

Often, the game-…