Artificial Intelligence
Artificial Intelligence
In the last 200 hundred years, mankind has made astronomical steps towards the future of modern technology and now we are fully plunging into the future of Artificial Intelligence (A.I). According to Merriam-Webster’s Dictionary, A.I is “a branch of computer science dealing with the simulation of intelligent behavior in computers,” (Forbes). The term first became popular in the 1950’s where computers were programed to solve mathematical problems and follow symbolic methods (SAS). As the government began to explore the future of A.I, the Hollywood scene started to fantasize about the face of A.I, which seemed to have inspired companies like IBM and DeepMind Technologies to create the likes of their own potentially terrifying robots capable of unfathomable destruction. Although something like Artificial Super Intelligence is still years away, the road to it is actually a lot closer than we think. An American engineer by the name of Gordon Moore predicted that the number of transistors per silicon chip doubles every year, in simpler terms we can assume that once we have achieved artificial general intelligence, then we will reach artificial super intelligence A LOT sooner than we think (Britannica).
As of today, we have mastered the concept of artificial narrow intelligence (ANI), so things like Google Maps, Self-Driving Cars, spam filters, google translate, the list goes on. A program like R and R Studio is an example of ANI because it is programmed to compute different statistical methods and problems. R studio cannot learn to suggest certain equations or databases, it essentially cannot better itself and learn in such a way like humans do. The R program is simply a calculator on steroids that is programmed to compute certain functions and create an output for them. It can calculate advanced data analytics that can tell us why exactly something is happening and help us predict what will happen in the future, however it cannot learn behaviors and retain knowledge. DeepMind Technologies have actually developed an algorithm that can learn to do just about anything, using differentiable neural computing (DNC) which simply allows the program to retain what is has learned (Future Timeline). Some of these AGI robots actually can learn from humans, which can be a very bad thing. A few years ago, Microsoft developed a AGI chat-bot that could learn human behavior via Twitter. In a matter of a couple of hours, the chat-bot went from nice and innocent to a “Nazi-Loving and feminist-hating racist,” (The Root). This goes to show, that maybe we do not want AI robots and programs to learn from human behavior, because then we are giving an already powerful robot, the hate resides in human nature.