How worried should you be about the robot takeover depicted in Terminator: Salvation?

How worried should you be about the robot takeover depicted in Terminator: Salvation?

How worried should you be about the robot takeover depicted in Terminator: Salvation?

What really happened.
May 21 2009 3:50 PM

Gaming the Robot Revolution

A military technology expert weighs in on Terminator: Salvation.

Terminator: Salvation. Click image to expand.
Terminator: Salvation

Terminator: Salvation, the fourth installment of the Terminator franchise, takes place in 2018, a number of years after an artificial-intelligence network devised by the U.S. military, called Skynet, has turned on its masters and set off a nuclear war. The sentient computer now controls an army of killer robots tasked with hunting down and killing every last member of the human race. It's Hollywood popcorn at its best, if not for the scary fact that the movie touches on some very real trends in modern military technology.

At about the same time that producers started working out how to make a Terminator movie without the governor of California, I set out to study what was happening as the machines of science fiction started to be used on our real-world battlefields. I interviewed scientists, generals, insurgents, and human rights activists for my book Wired for War and along the way discovered that our use of robotics in war is already surprisingly extensive. The U.S. military went into Iraq in 2003 with only a handful of unmanned systems in the air. We now have more than 7,000 drones in the force, including the famed Predator drone now gaining notoriety for its near-daily visits to Pakistan. Similarly, the invasion force that went into Iraq used no unmanned ground vehicles. The military now has more than 12,000 of them—including the lawnmower-sized Packbot, which is made by the group behind the Roomba robot vacuum cleaner but hunts for roadside bombs instead of dust bunnies. The next generation of machines now at the prototype stage are being armed with everything from machine guns to rockets.


There are hundreds, if not thousands, of American soldiers who are alive today because of robotics, so in many ways this triumph of technology should be celebrated. Many of the people I talked with described robotics as being equivalent to the rise of the steam engine or the computer in terms of its massive military, political, social, and economic ripple effects, making the Predators and Packbots of today akin to the Wright brothers' Flyer or the Model T Ford. Bill Gates described robots as being parallel to where computers were in 1980 and said we're "on the verge of a new era." But many others describe robots in the same way we now discuss the atomic bomb—as an invention we might one day wish we could take back. According to military robots pioneer Robert Finkelstein, the new technology "could end up causing the end of humanity, or it could end war forever."

This is important stuff that necessitates a debate in the here and now. But in our world, it is often pop culture that reigns supreme. So Slatehas asked me, a guy with a Harvard Ph.D. in security studies and who helped write the new president's defense policy agenda, to answer whether Terminator's robot revolt could ever come true. Fortunately, I am also an admitted sci-fi geek, so I am somewhat well-equipped to wrestle with the question of whether the metal ones might come for you.

Essentially, four conditions would have to be met before a Skynet-like entity could take over the world.

First, the machines would have to have some sort of survival instinct or will to power. In the Terminator movies, for instance, Skynet decides to launch a nuclear holocaust against humans in self-defense after the frightened generals attempt to take it offline. Yet most of the focus in military robotics today is to use technology as a substitute for human risk and loss. We use the Packbot in Iraq because, as one U.S. military officer tells, "When a robot dies, you don't have to write a letter to its mother." It would serve the very opposite goal to give our robots any survival instinct.

Second, the machines would have to be more intelligent than humans but have no positive human qualities (such as empathy or ethics). This kind of intellectual advancement may be possible—eventually—given the multiplicative rate at which computer technology progresses. But an explosion of artificial intelligence that surpasses humanity (sometimes referred to as the Singularity) is by no means certain. My Roomba vacuum, for example, still can't reason its way out of being stuck under my sofa, let alone plot my demise. There's also an entire field, called "social robotics," devoted to giving thinking machines the sort of positive human qualities that would undermine an evil-robot scenario. Researchers at Hanson Robotics, for example, describe how their mission is to build robots that "will evolve into socially intelligent beings, capable of love and earning a place in the extended human family."