Robert B. Fisher
Division of Informatics
University of Edinburgh
Robots and other artificially intelligent (AI) mechanisms are now a common plot device or even main character in movies [1, 2, 5, 6, 7, 8]. What is less well known is that they have appeared since the early days of cinema (1907). These movies are interesting, because they help shape the mainstream public's view of artificial intelligence and robotics. The experienced science fiction reader and AI professional have more developed views about AI, but these are minority of the population. Here I'm interested in how AI is understood and interpreted by the majority.
Continuing scientific developments are already bringing AI research results to the public: web search engines, network routing, automatic scheduling programs, automobile route planning, speech understanding, factory automation, home robot vacuum cleaners and lawn mowers, etc. But these seem like normal products and are not usually viewed as Artificial Intelligence, at least as AI appears in the cinema. The cinematic sort of AI is the key opinion former of AI: a challenge to humans because of their potential powers. This essay considers why these powerful cinematic AI agents, in spite of their presumed abilities, have largely abnormal ``personalities''.
A recent www-based survey of 256 movies containing elements of AI  has 83 ``true'' AI entries. From the 256 movies, I have excluded TV episodes (too numerous), short cartoons (too numerous), cyborgs (36), androids (20) and still unclassified movies (117). A debate could be held around the excluded cyborgs and androids, however, my concern here is with true Artificial Intelligence: the product of a non-human sensing and reasoning mechanism. Cyborgs have human brains (although perhaps physically or computationally augmented), so do not say anything about AI. Androids are more ambiguous as they are constructed entities. They are generally ``grown'' to be either identical to humans or ``improved'' versions. However, their mental machinery is assumed to be identical to humans, and so the use of an android says nothing about the implications of having an artificial reasoning mechanism. I have excluded these cases as also being uninteresting since they say nothing about the nature of non-human thought.
Of the 83 ``true AI" movies, 46 depict a ``mindless" sort of AI without self-reflection or self-awareness, largely as efficient robotic killing machines. Although this is an exciting plot device and a worrisome possible future, these AI agents do not have much depth and ultimately will interest viewers at about the same level as a well-engineered automobile. (Movies with mindless AI agents are listed in Appendix A.)
But what of the other 37 movies (listed in Appendix B)? Examples of these are Metropolis (1926), The Wizard of Oz (1938), 2001: A Space Odyssey (1968), Demon Seed (1977), Making Mr. Right (1987), Virtuosity (1997) and AI Artificial Intelligence (2001). These movies have AI-based agents as central characters, interacting with humans in many of the same ways as humans. They may not always be a cinematic or commercial success, but, because they explore the nature of AI agents, how the agents minds may work and the consequences of the integration of the AI agents into human society, they provide a glimpse at how the rest of society views AI agents. This view is not a healthy one.