SIGGRAPH was great this year. Highlights from the Art Gallery included interactive pieces such as Brad Paley’s TextArc (which I’ve been meaning to write about here) and papers such as Gonzalo Frasca’s (which contained the great line “Representation is degree zero of simulation”).

Unfortunately, next year’s SIGGRAPH Art Gallery chair is not planning to have interactive art or Art and Culture Papers.

Elsewhere at SIGGRAPH, Andrew Stern and I organized a panel in the main program which went quite well - if I do say so myself - despite some changes in panelists that will make anything you’ve read in the SIGGRAPH program or proceedings innacurate. The panel was titled “Interactive Stories: Real Systems, Three Solutions” and presented perspectives that might be called “narrative game,” “interactive drama,” and “story recognition.”

our first presentation was from James Leach, the author of the story and dialogue for Black and White (replacing original panelist Peter Molyneaux, the designer of Black and White, who had been asked by the EC to meet with them about European game policy). James spoke about B&W from the point of view of a writer - talking about how the game’s events were written and specified in such a manner that the setting, actors, and lines spoken for each could be adjusted, as appropriate, to what had taken place so far. Black and White contains more than 60,000 lines of dialogue, each of which has emotion and animation directions embedded in simple square-bracket cues. James talked about B&W in the context of previous projects (he was also in-house writer at Bullfrog, where he worked with Molyneaux on projects such as the Dungeon Keeper games and Populous) and called it their first serious attempt at interactive story - as opposed a multiple-path, linear progression, sandbox, or other type of experience. I had a great time talking with James about writing for electronic media, and I’m hoping to go visit him at Lionhead next summer.

The next presentation was from Andrew and Michael Mateas, who are collaborating on an incredibly ambitious project called Facade (which was presented in a much more embryonic form at DAC 2001). This project combines significant advances in drama management and interactive characters with project-specific techniques for animation and natural language, creating a platform for an interactive, first person, one act drama. I say “platform” because the technical system, however accomplished, is just a place for good writing and animation to be built up. In the case of Facade, this material tells the story of a marriage that is dissolving before your eyes - resulting in a sort of interative “Who’s Afraid of Virginia Woolf?” or, as Nick Montfort puts it, a first person shooter with a cocktail glass. Response to this presentation was amazing. Not only did the audience laugh and exclaim during the video shown during the talk, they also applauded at the video’s end (not common at SIGGRAPH), applauded again at another video clip shown during the question and answer period, and then were heard to be repeating lines from Facade to each other in the lobby after the panel was over.

The last presentation was from Will Wright, designer of groundbreaking projects from the 1980s SimCity to, more recently, The Sims (replacing original panelist Stuart Moulthrop (Victory Garden) who had earlier been replaced by Bernie Yee (EverQuest) who had to withdraw because of a personal emergency). Will talked about how The Sims, a project he developed as a simulation game, is being used for storytelling by many players. Given this, he talked about his plans for beginning to support story in future versions of The Sims - to a limited extent in the upcoming online version, and to a greater extent in an unspecified future single-player version. Will doesn’t want to do this in anything like the traditional interactive story modes (e.g., rail-like adventure games) but rather through what he calls “story recognition.” A story recognition approach, instead of trying to tell a pre-written story or offer users a choose-your-own-adventure, pays attention to what users are doing and then attempts to support their activity. For example, if the user seems to be acting out a type of genre story, the system could begin to play appropriate music or make available appropriate props (if it seems like a horror story, a chainsaw might be there when the character enters the next room). If the player seems to be telling a story of uncertain or mixed genre, a selection of props might be offered (if it seems like it could be horror, comedy, or horror/comedy, a chainsaw and creampie might both be in the next room). Similarly, if the player seems to be telling a story about reaching a particular goal, the system could begin to offer appropriate challenges to reaching that goal. At a conference where many seemed to be saying, “Games need better AI” without any clear idea of what would be done with this better AI, Will’s presentation was not only intellectually exciting, but also a breath of fresh air.

After our panel came “Games: The Dominant Medium of the Future?” - an excellent panel, though also upsetting, because moderator Ken Perlin told us that the rumors we’d heard were true: the executive committee for next year’s SIGGRAPH has decided to cut panels from the conference program. The reason given, as with the rumored cuts to the Art program, is budget difficulties. But I believe it’s a big mistake. I know I won’t attend a SIGGRAPH with no panels, no art papers, and no interactive art. And I know others for whom these programs, singly or in combination, are primary reasons for their SIGGRAPH attendance. Cutting these programs entirely, rather than paring them and other programs back more incrementally, will only lower SIGGRAPH attendance further next year, and exacerbate SIGGRAPH’s financial problems (which began with the lowered attendance all ACM conferences have been experiencing). If you care about SIGGRAPH panels, I urge you to write to the conference chair (chair-s2003@siggraph.org) as soon as possible. And if you care about SIGGRAPH presenting art papers and interactive art, I urge you to write to next year’s art chair (art-s2003@siggraph.org).

If SIGGRAPH shoots itself in the foot by killing the programs that many attendees care about most then it is unlikely the money will be available later to restore them. The programs won’t come back, the attendees won’t come back, and SIGGRAPH will become a narrow conference rather than the exciting interdisciplinary environment it has been. It was a great SIGGRAPH, but I’m worried it may have been my last.