From the fantasy of science fiction to the labs of science, the disruptive technology of Artificial Intelligence (AI) may soon see man lose against machine.

During Zeitgeist 2015 in London, renowned physicist Stephen Hawking warned that computers would control humans in the near future:

“Computers will overtake humans with AI at some [point] within the next 100 years. When that happens, we need to make sure the computers have goals aligned with ours”.

A hero in a Terminator-themed narrative could have uttered this dramatic warning. However, Hawking’s urging of an AI-alliance has an increasingly jarring ring of truth to it. The implication that humankind could be dominated by its own technologies is rooted in everyday examples of Machine Learning (ML).

Artificial intelligence for lean, mean, thinking machines

Experts generally regard ML as a sub-sector of AI. It is most noticeably present in what have become mundane details of daily life.

Algorithms drawn from user behavior can generate content recommendations in platforms such as Netflix or Spotify. Brands use data from purchasing habits to personalize advertising. Furthermore, websites such as Facebook, Reddit, Google, Twitter and Tumblr have protected algorithms to rank content.

The core premise of ML is to enable machines to learn for themselves from data. Computers and similar technologies are currently being coded to think like people. This goal is pursued through the development of artificial neural networks inspired by those native to the human brain. These machines then develop their fledgling cognition through exposure to an almost infinite body of digital information: the Internet.

ML tools represent an ability to learn through data collection. However, they are incapable of the mental gymnastics that characterize the human experience. Self-awareness, creative thinking, raw curiosity, and idle daydreaming are complex cognitive processes that divide man from machine.

This indulgent hubris is tempered by the evolution of a subset of ML, Deep Learning (DL), which takes the premise of neural networks even further. Engineers are creating sophisticated algorithms that mimic complex human thought and abstractions in the hopes of developing artificial intelligence that rivals that of mankind.

With self-driving cars and increasingly automated facets of life fast approaching, the apocalyptic fear of advanced computerized processes may not include the gloom and doom held by Hawking.

However, in light of the swift technological developments sweeping through the world, and the degree to which humans depend on their computers and devices, perhaps a computer contingency plan would come in handy in the next century.

Read more:

The Steve Jobs leadership style can teach you a thing or two