I may not be a good actor, but that's ok. Being an Assistant Professor of Computer Science at the University of Chicago doesn't require much acting. I've acted just once, in the film Computer Chess, premiering this week at Sundance. Director Andrew Bujalski's fourth film is set circa 1980 at a computer chess tournament in which competing teams play their chess-playing computer software and hardware against each other. Bujalski modeled the film's setting on the ACM (Association for Computing Machinery) Computer Chess tournaments that ran from 1971 through 1994, peaking in intensity and popularity around 1980.
By choosing this particular setting at this particular time, Bujalski puts his characters in a time of heady ideas about the promise of artificial intelligence, when our modern understanding of the scope and capabilities of computers was taking shape. With computing power and programming sophistication at a fraction of what they are now, in 1980 the question of whether a computer could beat a chess Grandmaster was still open. That was answered with the victory of IBM's Deep Blue computer over Garry Kasparov in a well-publicized 1997 match. In hindsight, we know that the more interesting question is not whether, but when and how, a computer would beat a Grandmaster. The ingredients of Deep Blue's successful design were forged in the computer chess tournaments that Bujalski's film visits.
Computer chess tournaments were a vital incubator and testing ground for new ideas in computer chess specifically and in artificial intelligence (AI) more generally. Leaders of AI such as Allen Newell and John McCarthy worked on computer chess, and popular press about the early computer chess tournaments treated them as a yardstick of computers' thinking ability. There were also philosophical questions in the air, such as whether a program was capable of creative or insightful play, and to what extent we could learn from the computer when it played a strong game. Brilliant and passionate people, with intellectual goals as well as complex emotional investments, populated all sides of these issues. Bujalski's characters live in this moment, working through the ideas and ambitions of that optimistic and fertile time.
As reflected in Computer Chess, it was also a time of flux for what it meant to be a computer and how we used computers. Large mainframes in remote machine rooms were giving way to smaller minicomputers sitting in offices, which led to desktop personal computers. Bujalski tracked down a range of period-appropriate computing technology with the help of the Goodwill Computer Museum in Austin TX, which loaned their teletype terminals, old acoustic-coupler modems, various minicomputers, and a few clunky PCs. Some of these machines were coaxed into running again by volunteers at the museum, who in turn became extras in the film. Bujalski also enlisted the consulting expertise of Peter Kappler, a software developer who has both created chess programs and played in (human) chess tournaments. Kappler devised interesting positions and moves for the games shown in the film, and he was on set during filming to help keep the dialog about hardware and programming consistent with the period.
As for me, my role in the film was that of actor and consultant. Even though I'm in computer science, my area is scientific visualization and image analysis, conceptually far from the discrete mathematics and decision trees required for computer chess. But to get into character, and to avoid creating a permanent record of my ignorance, I did some research into this field, including talking with David Slate, a tournament-winning chess programmer in the 1970s and 1980s. His stories of imagining, writing, and debugging complex programs on uncooperative computers were jarringly similar to the challenges I face developing software for my own work. To this day, computers are only as smart as the software running on them, and creating good software is a human effort, requiring creative vision and dogged persistence.
By the end of the 1980s (after when Computer Chess is set), winning programs usually used brute computing force to evaluate future moves instead of mimicking the selective search of human players, and the software ran on bespoke hardware specifically optimized for chess. There is in fact an uninterrupted line in personnel and technology from the brute-force programs created for the computer chess tournaments, to the team at IBM that created Deep Blue to beat Kasparov. Along the way, the AI research community created foundational methods of efficiently navigating the space of moves and countermoves in games like chess. They also learned that there is much more to mimicking human thought than winning games. But then, what tasks should we use as indicators of successful artificial intelligence? What have we learned about ourselves when computers start matching our abilities? How do we manage our relationship to computing technology as it shifts and grows around us? Bujalski's film offers a glimpse into a time when we were learning how to ask those questions.
I'll be seeing the film this week along with my wife, who plays my wife in the film, and our infant daughter, who plays our infant daughter. The person in our family who stretched their acting abilities furthest was me, since I'm an Assistant Professor in real life, not a famous Professor with an empire of graduate students pursuing my research agenda. As for my acting, I'm not quitting my day job, unless my tenure decision goes badly. In which case, Mr. Spielberg, you know where to find me.