Artificial Intelligence in Science Fiction

Jason Ellis: Fall 2004

 

 

Karl Capek - R.U.R. (Rossum's Universal Robots)

Karl Capek's 1921 play, R.U.R. (Rossum's Universal Robots) is an example of an analog dystopic AI. This work introduces the term "robot" to the English language, but the Robots (Capek's capitalization) in R.U.R. are more like androids than robots. The Robots are shaped like humans, but the character Domin says that they are made "from a different matter than we are." These Robots have perfect memories but they are not self-aware. Memory is divorced from self-analysis. Using industrial chemical processes, the Robots' individual pieces (arms, legs, organs, etc.) are cooked up from "batter" in "kneading troughs" and "mixing vats." Then, those components are mated into a whole Robot in an assembly line operation. Thus, gears and cogs are not present in Capek's Robots, but the means of its creation are partially mechanical as well as chemical.

The leaders of R.U.R. are attempting to create a utopia for humanity by pushing off the drudgery of work onto the many Robots that it creates. Dr. Gall, who is in charge of the "physiological and research divisions of R.U.R.," modifies a few robots to be more human-like, and in doing so, "they stopped being machines." These modified Robots incite the other robots to destroy all of humanity, their collective oppressor. After all of the humans save one are destroyed, the Robots begin to fear death. The last human, Alquist, who is the constructor of R.U.R., is told by his captors to rediscover the lost science of creating Robots. Ultimately it doesn't matter that Alquist fails. When he witnesses the beginning of love between two modified Robots, Helena and Primus, he exclaims, "Now let Thy servant depart in peace O Lord, for my eyes have beheld...Thy deliverance through love, and life shall not perish!" It doesn't matter that Alquist is unable to build new Robots because somehow things have changed (either through Dr. Gall's undisclosed modifications or through some other process) so that the Robots are capable of being human (e.g., feeling emotions of love, fear of death, and being able to procreate).

 TOP

Isaac Asimov - I, Robot

Isaac Asimov's short story collection, I, Robot (originally published by Gnome Press, 1950) is primarily representative of digital utopic AI. The collection contains nine of Asimov's early robot stories. The stories are tied together as an interview with the retiring robopsychologist, Dr. Susan Calvin. She is the best choice for this narrative because she is there from the beginning, literally. She is born in the same year that U.S. Robots and Mechanical Men, Inc. is founded and later, after she obtains her Ph.D. she is hired by U.S. Robots as a "'Robopsychologist,' becoming the first great practitioner of a new science" (I, Robot xii). She bridges the physical sciences with the science of the (robot) mind. Also, all of the stories are linked by Asimov's Three Laws of Robotics which are supposed to control the way that a robot reacts and reasons. These Laws, as listed in the short story "Runaround," dictate that:

1) A robot may not injure a human being, or, through inaction, allow a human being to come to harm.

2) A robot must obey the orders given it by human beings except where such orders would conflict with the First Law.

3) A robot must protect its own existence as long as such protection does not conflict with the First or Second Laws.

A strong example of digital utopic AI appears in "Evidence." This story introduces Stephen Byerley, who is running for the mayor's office. The problem is that his opponent believes that he is a robot. The circumstantial evidence points to the possibility of Byerley being a robot, but even if he is, then he would be the best person for the job because by following the Three Laws he would be the perfect caretaker for his constituency.

Most of the stories in I, Robot are utopic because the robots are depicted as being humanity's helpers and caretakers, but there is one dystopic story, "Little Lost Robot," in which a Nestor robot tries to run away and, when he is discovered, to kill Dr. Calvin. Asimov's carefully crafted Three Laws provide stability in robots' positronic brains. The Nestor robot featured in this story has a shortened version of the First Law which is stated as, "No robot may harm a human being" (I, Robot 143). The weakened First Law allows this robot to develop a superiority complex, which leads to its attempt to kill Dr. Calvin when she discovers him. Thus Asimov uses even his dystopic robot stories to demonstrate the significance of a robot's programming upon its relationship to humanity.

 TOP

Arthur C. Clarke - 2001: A Space Odyssey

Arthur C. Clarke's 2001: A Space Odyssey is an example of a digital dystopic AI story. A select few humans learn that mankind is not alone in the universe after an alien artifact (the Monolith) is discovered buried under the surface of the moon. When the Monolith is exposed to the Sun, it emits a brief, but intense radio signal that is directed toward Japetus, one of Saturn's moons. The spacecraft, Discovery, is sent to Japetus carrying one AI and five humans. The AI is a HAL 9000 computer system, known simply as Hal. Of the five humans aboard Discovery, three are in hibernation. The two who are awake, Dave Bowman and Frank Poole, maintain the ship with Hal. Eventually, a conflict develops in Hal's "subconscious" because it cannot reveal the true nature of the Discovery's mission to Bowman and Poole. This leads Hal to make mistakes that Bowman and Poole interpret as threats on their lives. After Hal kills Poole, Bowman chooses to "disconnect" (i.e., kill) Hal in order to regain control of the ship. Bowman goes on to Japetus where he finds a larger Monolith. This Monolith is actually a "Star Gate" that transports him far from our solar system. When Dave reaches his final destination, the aliens transform him into a being without physicality, but as a child with eons before it in which to grow.

Although the story as a whole addresses human evolution, the sequence with Hal is both the longest and most gripping, demonstrating Clarke's specific interest in the similarities between human and machine evolution. Evolution manifests itself through human and machine programming. The monolith programs early humans and modern humans program Hal. Hal appears to be crazy and intent on murdering his crewmates. This is why Bowman chooses to disconnect him. However, Hal is an AI whose identity is built on software and hardware that is too complex for one person to comprehend the whole system. There is a reason to his madness and no reasonable amount of prior testing might have elicited Hal's behavior aboard the Discovery. He was ordained with priorities and mission objectives that acted as a program that must be run to completion because that is what computers do--run programs. Because Hal's "mind" is modeled after the human mind, the symptoms and actions that Hal exhibits are similar to the way in which a neurotic human might act. Despite what Hal has done we feel sorry for him by the end because, like humans, he fears death.

 TOP

William Gibson's - Neuromancer

William Gibson's 1984 novel, Neuromancer is a more recent example of digital dystopic AI and a prime example of the cyberpunk movement in SF. The story is set in Earth's future where an AI called Wintermute who has a compulsion to connect/merge with another AI called Neuromancer. Wintermute orchestrates his liberation by bringing together several carefully chosen humans who can beat the failsafe that keeps him caged in the Berne AI mainframe. Case, the net cowboy, works with a construct and a military grade virus to break through the ICE security around the Berne AI mainframe. Molly is a razor girl who protects Case and she interacts with the physical world while Case jacks into the matrix. Armitage serves as a physical presence for Wintermute in the same way that a computer construct in the matrix works on behalf of a human operator. After the ICE is broken with the help of Case's associates, Wintermute is able to merge with Neuromancer to become an entity greater than anyone could have imagined.

The story involves several instances of AI designed by humans for human ends. The lowest form of AI is the Braun, a small spider like work robot that Wintermute uses to guide Molly and Case inside the Villa Straylight. One of the highest forms is the construct, Dixie Flatline. A construct is a limited form of AI based on the memories and experiences of a dead human being, in this case the famous hacker, McCoy Pauley. The two primary examples of course, are the strong AIs present in Wintermute and Neuromancer. Wintermute is a calculating AI that is explicit in its manipulations. Neuromancer is more personality based and he uses subtle manipulation. Wintermute is located in hardware in Berne while Neuromancer is running on hardware in Rio. These two AI entities are two halves of one whole. The mega-corporation, Tessier-Ashpool, which gave birth to these AIs, had them separated with safeguards imposed by the Turing police. They both have limited citizenships as individuals because of their self-awareness, but the extent of their knowing and understanding has been limited due to the division. As the reader learns, Marie-France, the matriarch of the Tessier-Ashpool clan, probably implanted the drive within Wintermute to break free and unite with his "brother," Neuromancer. Not surprisingly, these AIs use the products of capitalism (e.g., hiring "mercenaries" and using information as power over others) to shuck their chains binding them to Tessier-Ashpool. Thus, the AIs use human beings for AI ends.

 

   

TOP