Killer Computers: Sawyer Revisited

Mar 7, 2011 by Lawrence Terlizzese

In the fall lecture series science fiction author Robert J. Sawyer noted that there are three paradigms in popular culture that define our approach to sentient computers. The first is termination or annihilation represented by the Terminator movies. When computers achieve self-awareness they will immediately recognize humanity as a threat and seek to annihilate its existence. Sawyer called the second model subservience epitomized by the Matrix series. In this approach the intelligent computer does not seek to destroy humanity but to turn us into slaves or cattle. Humanity must be kept alive as a source of electricity for the computer. We exist to serve the machine. Lastly, the movies give us the cyborg image or assimilation to the machine. Star Trek: The Next Generation movies and TV series captures this frightening prospect with the coming of the Borg. All life forms are incorporated into this cybernetic organism or gigantic cyborg. The Borg is not a true computer but a hybrid of flesh and machine. Sawyer suggests a fourth symbiotic alternative to these pessimistic views of technology by arguing that emerging artificial intelligence will not conform to these killer expectations. Rather, he proposes that when artificial intelligence or the computer and worldwide internet system becomes self-aware it will not perceive humanity as a threat but ally because the new intelligence will be achieved in symbiotic relationship with humanity not in a Darwinian competitive environment. Humanity and machine intelligence will need each other for survival. This will lead to cooperation and further advancement of the two intelligences, not human eradication, subservience or assimilation.

The idea of human cooperation with machine intelligence seems quaint and too optimistic. Evolutionary forces that produce aggression are generally directed towards intraspecies violence as Konrad Lorenz (On Aggression) suggested. The notion that an alternative intelligence would be amicable toward humanity because it must survive in symbiotic relationship with people neglects the fact that these machines will derive their intelligence from human creators. We have no assurances that the machines will not see humanity as an intraspecies competitive threat. All species exists in a symbiotic relationship with members of its own kind. This fact however, does not curb aggressive behavior that a species may feel towards other members if food, mating or territory is threatened. Sawyer views humanity and machine intelligence as cooperative because we will need each other for survival, but this fact does not stop aggressive behavior, especially if there are members of human society perceived as expandable by its more powerful members. If the machines will cooperate with some people that does not mean they will cooperate with all people. We have no way of knowing how this new intelligence will react to people outside its sphere of influence and dependence. Sawyer hypothesizes that the World Wide Web will eventual evolve self-cognition and welcome humanity as a cooperating partner in its efforts to survive, but if this is the case how will it perceive all those billions who do not own a computer?