How fast will it take before robots turn on humans?
Science Fiction writers have warned us for decades about trusting robots. Hal 9000, the onboard computer for Space Odyssey 2001, was also considered a friend and companion. Hal functioned as the sixth crewmember and was responsible for life support systems, guidance, and other ship functions. Hal was an all around good guy until the day he decided to murder several of the team members.
Master Control Program, the antagonist in TRON, had designs on world domination. Skynet is a computer system that has haunted American minds since the 1980’s. The Global Digital Defense Network, designed to make decisions regarding national defense, chose first to protect itself from the destruction of mankind. By eliminating the fallibility of human decision, Skynet thus sought to ensure it’s own survival.
Modern military robots
The modern military is experimenting with autonomous weapons systems (AWS), self-targeting, self-executing defense systems. Robotic weapons have no conscience, no fear, and no morals, thus making kill decisions on program parameters. Perhaps the addition of AI will allow the AWS to learn and better determine what is a threat?
And yet, consider our opening question of “How fast will it take before robots turn on humans?” Just how fast will it take the AWS to turn on their makers? According to recent tests by Microsoft Corporation, about two days. Microsoft turned their chat bot, known as Tay, loose on Twitter to learn to talk like a millennial, and after only two days of interaction Tay was “racist, homophobic, and all-around bigoted.”