April 2017

40 years ago today, science fiction’s most iconic film was released. Star Wars gave us two of our most beloved onscreen robots. Though R2-D2 and C3P-O were far more advanced than anything existing then, and even now, artificial intelligence was already well on its way when the film premiered. In fact, the father of AI had already designed a test to determine a machine’s intelligence back in 1950. The Turing Test, created by mathematician and computer scientist Alan Turing, asked a judge to field a series of questions to two participants: one machine and one human. Based on the responses, the judge would then determine the human responder. The computer was deemed intelligent if it was able to fool the judge at least 50 percent of the time.

However, what exactly is artificial intelligence? Can we expect benevolent robots to be walking amongst us in the near future? Or should we be more concerned about AI warping into a more sinister HAL 9000? 

Simply put, Artificial Intelligence (AI) refers to a machine’s ability to imitate human cognitive abilities like problem solving and learning. These skills also require an understanding of language and speech and strategic thinking. We already encounter numerous applications of artificial intelligence on a daily basis: when we ask Siri for directions, when driving past Google’s autonomous vehicles, when we use self-parking or cruise control features. 

  
However, inspired by film and literature, many people are more familiar with the concept of

Artificial General Intelligence (AGI). AGI computers would be equally intelligent to humans in every aspect. These machines would be capable of performing the same kind of intellectual tasks that humans perform with the same level of success. Apple’s co-founder Steve Wozniak developed the Coffee Test as an indicator for AGI. During the test, a robot has to enter an average home and attempt to make coffee. This means finding all the tools, figuring out how they operate and then performing the task. A robot able to complete this test would be considered an example of AGI. Think C3P-O!

But what about HAL 9000? The 2001: A Space Odyssey, horror computer is also still currently the stuff of fiction. This field of AI is called Artificial Superintelligence (ASI), and it is also the field of research that Stephen Hawking has warned us about. One of AI’s leading figures, Nick Bostrom has defined superintelligence as “an intellect that is much smarter than the best human brains in practically every field, including scientific creativity, general wisdom and social skills.” This slight or vast superiority to human intellect is disconcerting to scientists and laymen because its outcome is unknown. A machine capable of constantly learning and improving itself could be unstoppable. 

Where exactly is AI headed? Some of the most exciting advancements in AI are a result of self-learning and analytics. For example:

• A study published by Nature outlined the work of a team of researchers on “simulated childhood”. This is a robot’s ability to detect a malfunction and use a process of trial and error to determine the defect and repair it. In a matter of minutes, the robots tested could identify a problem and determine the best ways to fix itself. The robots could repair 14 forms of damage.
• In 2013, a computer had the intelligence level of a 4-year old child. In 2014, a computer was able to solve a mathematical problem so complex that the most intelligent human minds could not check for accuracy. Computer scientist and futurist Ray Kurzweil said that computer intelligence would reach human levels by 2029. 

• By analyzing data from over 100 million articles, SGI Altix supercomputer Nautilus was able to predict the future. The computer could assess tone and mood, connect it to geographical locations and use that information to foresee, for example, Arab Spring and Osama Bin Laden’s secret location. Internet entrepreneur Kalev Leetaru ,who fed the computer all the information, considers it akin to weather forecasting. “It’s never perfect, but we can do better than random guessing.”

It is clear that AI has resulted in machines that are, in many ways, more capable than human minds. However, it will be a long time before they are able to surpass human intelligence in every regard.  And should this day come, the brightest minds in the field are well aware of the risks of superintelligent computers and should be able to thwart any HAL 9000’s before it’s too late!