By Lucas Wood
Artificial intelligence has its roots in both myth and science. For instance, the Greek god Zeus created Talos, a brass robot. Talos was designed to protect the human woman with whom Zeus was currently smitten, Europa, the mother of the King of Crete. Talos was tasked with circling Crete three times daily to guard against pirates and other invaders.
Creation myths were not limited to the Greeks. Twelfth-century Jewish folklore mentions golems, anthropomorphic creatures created from clay or mud by virtue of a word etched onto their forehead. Fifteenth-century writings on alchemy spoke of the creation of a fully formed human (Homunculus) in miniature. These works and others spoke of mechanical or otherwise inanimate creatures granted the ability to reason and the will to act independently by their creators.
In the 20th Century, Artificial Intelligence Begins to Approach Reality
In the 20th century, artificial intelligence started its transformation from ancient myth and legend to manmade reality. The Turing Machine and discoveries like electrical and neural networks led scientists to validate what philosophers like Aristotle, Hobbes and Descartes had concluded: that logic can be distilled into formal methods of deduction to create a “thinking machine.”
In 1956, prominent scientists and mathematicians converged at Dartmouth College and proposed “that every aspect of learning or any other feature of intelligence can in principle be so precisely described that a machine can be made to simulate it.” The term “artificial intelligence” (AI) was born.
But history is also filled with cautionary tales warning against the potential for disaster at the hands of the machines we create. In the 1940s, science fiction author Isaac Asimov popularized three cardinal laws to govern “robots” (as he described them), a decade prior to the Dartmouth conference.
These laws were created in response to an old science fiction story in which creations of artificial intelligence turned on and destroyed their creators. From the 1940s onward, popular science fiction has been transfixed with the advent of true artificial intelligence and the so-called “rise of the machines.”
Modern Artificial Intelligence Benefits Us in Multiple Ways
Today, we are realizing the benefits of more rudimentary AI across society. From welding robots in auto manufacturing to robotics competitions, we are creating limited-autonomy robots capable of completing brief tasks without human control or intervention.
Another contemporary term akin to AI is ‘machine learning,’ which is an advancement that many cyberdefense technologies now boast as having. Digital assistants such as Apple’s Siri and Microsoft’s Cortana, in a bid to put a near-human voice on the face of machine learning, have attempted to capitalize on the vast resources of information on the Internet and the near ubiquity of processing power in consumer electronics.
How Long Will It Be before AI Becomes Sentient?
While artificial intelligence has yet to progress to the point of being sentient (the ability to be, feel and suffer), the depth and power of artificial intelligence grows with increasing speed. With the introduction of new technologies come moments of human fear as the technologies we create appear to take on traits long heralded as the hallmark of sentience.
For instance, Facebook’s recent display catalyzed these human fears when its “chat-bot” prototypes began communicating with Facebook users in verbal shorthand. Observers declared these prototypes as signs of uncontrolled evolution and they were quickly labeled as alarmist by others. Artificial intelligence experts lined up to reinforce Facebook’s statement that the chat-bot prototypes were not a herald of the “Skynet” apocalypse, but rather a predictable and controlled behavior of AI.
Will AI Become Self-Creating?
So should we proceed cautiously toward an age when machines begin communicating and even creating themselves without human intervention? Some people might say we have no choice.
Mikko Hypponen, Chief Research Officer at F-Secure, recently gave a talk at a Black Hat 2017 information security conference about the next 20 years in cybersecurity. He stated that we are entering an era when computer programs can begin creating other computer programs.
In the next two decades, artificial intelligence will not only become smarter and able to do more complex tasks. AI will also be able to create new programs and accomplish new, formerly un-programmable tasks by constructing its own tools. According to Live Science contributor Charles Q. Choi, this cognitive leap toward the creation of one’s own tools to accomplish complex tasks is “tremendously significant.”
Where Do We Go from Here in Artificial Intelligence?
There’s no doubt that we are proceeding at ever-quicker speeds toward autonomous artificial intelligence, a concept which both the renowned theoretical physicist and cosmologist Stephen Hawking and modern technology luminary Elon Musk warn may spell our demise. They have since devised a set of 23 principles to cope with this evolution, yet it is hard to know whether anyone is listening.
What we do know is that just as we begin to master technology to cure disease and improve healthcare, we are becoming just as productive in creating new ways to wage war. As these technologies evolve, so too must our understanding of the ethical ramifications of their autonomy. Arguments persist today, for example, on the development of autonomous weapons. Regional security organizations struggle to develop international norms to cope with the eventuality that machines will be gifted with guns and the intelligence to use them.
Whether or not you believe that machines are destined to confiscate the mantle of power and take control over human civilization, one thing is for certain. Now is the time to pay attention.