Irrespective of the success of the movie as an exciting, well-told story, or the plausibility of the AI agent, movies containing agents with `minds' evoke interesting questions based on comparisons with human behavior. From this perspective, there is a notable fact about almost all of the AI agents in these movies: in varying degrees, the agents show abnormal behavior, from obsessive to pathologically insane. The main forms of deviant behavior are:
These are all common human behavior disorders, when taken to extremes. Other types of mental abnormality that afflict humans in varying degrees could or very occasionally do appear in the movies:
Coming back to the cinema, these forms of abnormal behavior have had some but not a lot of exploration from the perspective of central characters in movies (e.g. Skynet in The Terminator is the cause of the main conflict but does not appear in the movie). So, we might see some more movies pursuing these aberrations.
Why then is insanity or extreme behavior so common?
People have always had anxieties arising from the uncertain future that a new technology might create. In this general sense, AI agents are no different than other technological devices, such as genetic engineering and nuclear power which also cause much justifiable anxiety. In those cases, the technology is already actively practiced, whereas AI technology is very far from being able to cause the problems that people worry about.
An AI agent makes our technological anxieties clearly visible, because the technology has the potential to displace not only our physical, but also our mental labor, and potentially even ourselves as the dominant agent on the planet. Given the success of mechanical devices, it is clear that people generally expect robotic devices to be stronger. Further, given the constantly increasing speed of computers, most general viewers would expect an AI agent to think faster than a human. Thus, having an AI agent in a movie allows us to explore our relationship with future AI agents, whether as master, equal or slave.
To give humans space to still be superior, movies credit humans with the ability to think better, clearer or more sensibly. Having an insane AI agent thus gives our poor human egos some boost, much like the `mad scientist' caricature expresses general anxieties about the power that comes from privileged knowledge and reassures the rest of us that ``at least we're still sane''. Alternatively, insanity gives AI agents a potential ``Achilles heel'' to exploit in conflicts. Hence we see AI agents with flawed reasoning of various sorts, such as not having real-world experience or common sense or not knowing something that only born, feeling, mortal humans would know. The technical reasons that lead to these shortcomings may be hard for the general public to understand. Insanity, on the other hand, is easily understandable and so can be more easily exploited in a story.
The scientific era has also subjected humans to an extended loss of status, largely leaving only our spiritual dignity intact (so far). People do debate whether animals possess consciousness and some spiritual side, but no mainstream religious or spiritual beliefs claim that mechanisms have souls. This implies that there should be a difference in behavior between humans and AI agents. Given the value placed on having a soul or spirit, this places AI agents in an inferior spiritual position. Although this does not imply insanity, it certainly implies ``not quite human''.