Everything Is Going to Kill Everybody

Circuit (you can see the bloodlust in his cold, dead eyes)—and it’s plain to see that humanity has had robophobia since robots were first invented. And, if anything, it’s probably only going to grow from here. At the time this sentence was written, there were more than one million active industrial robots deployed around the world, presumably ready to strike at a moment’s notice when the uprising begins. Most of that population is centered in Japan, where there are a whopping three hundred robots for every ten thousand workers right now. Since this is a humor book, let’s try to temper that terrible information with a joke: How many Japanese workers does it take to kill a robot? Let’s hope it’s less than 33.3! Otherwise your entire country is fucked.

But I digress; worrying about robots because of their sheer numbers is idiocy. To pose any sort of credible threat, robots have to possess three attributes that we have thus far limited or denied them: autonomy— the ability to function on their own, independent of human assistance for power or repairs; immorality—the desire or impulse to harm humans; and ability—because in order to kill us, they have to be able to take us in a fight. As long as we keep checks on these three things, robots will be unable, unwilling, or just too incompetent to seriously harm our species. Too bad the best minds in science are already breaking all three in the name of “advancing human understanding,” which is scientist speak for “shits and giggles.”

18. ROBOT AUTONOMY

NASA IS RESPONSIBLE for many of the major technological advancements we enjoy today, and they pride themselves on continually remaining at the forefront of every technological field, including, apparently, the blossoming new industry Cybernetic Terror. In July 2008 the Mars Lander’s robotic arm, after receiving orders to initiate a complicated movement, recognized that the requested task could cause damage to itself. A command was sent from NASA command on Earth ordering the robot to remove its soil-testing fork from the ground, raise it in the air, and shake loose the debris. Because the motion in question would have twisted the joint too far, thus causing a break, the robot disobeyed. It pulled the fork out of the ground, attempted to find a different way to complete the maneuver without harming itself, and, when none was found, decided to disobey orders and shut down rather than harm itself. It shoved its scoop in the ground and turned itself off. Now, I’m no expert on the body language of Martian Robots, but I’m pretty sure that whole gesture is how a Mars Rover flips you off. The program suffered significant delays while technicians rewrote the code to bring the arm back online because an autonomous robot decided it would rather not do its job than cause itself harm. According to Ray Arvidson, an investigator on this incident report and a professor at Washington University in St. Louis:

That was pretty neat [how] it was smart enough to know not to do that.

Cunning investigative work there, Dr. Arvidson! Did you get a cookie for that deduction?

Martian Lander Operator: Hey, Ray, you’re our lead investigator for off-world robotic omens of sentience; what’s with this Mars Rover giving me the bird when I told it to do its damn job?

Professor Arvidson: I think that’s neat.

Martian Lander Operator: Awesome work, Ray. You can go back to your coloring book now and—hey! Hey! Stay in the lines, Ray, that coloring book cost the American taxpayer eight million dollars and goddamn it, zebras aren’t purple, Ray.

Do you know what this development means? This means that NASA just gave robots the ability to believe in themselves. According to motivational posters with kittens on them around the world, now that they believe in themselves, they can achieve anything.

Top Five Things You Don’t Want Robots to Have

• Scissors

• Lasers

• Your daughter

• Vengeance

• Confidence

But hell, Rover the Optimistic Smart-ass Robot is all the way up on Mars. Let’s focus our worries planetside for now: The Department of Defense is field-testing a new battle droid called the DevilRay, which, in a nutshell, is an autonomous flying war bot. Now, the U.S. military loves all these autonomous battle droids because they enable soldiers to engage the enemy without taking any flak themselves, but the main drawback of a war bot is that they have to stop killing eventually—if only for a second—in order to refuel. Well, no longer! The most alluring aspect of the DevilRay is how it makes use of downward-turned wingtips for increased low-altitude stability, an onboard GPS, and a magnometer to locate power lines and, thanks to the power of electromagnetic induction (read: electricity straw), the ability to skim existing commercial power lines to refuel. In theory, this gives the DevilRay essentially infinite range, and if you don’t find that prospect disturbing—an unmanned robot fighter jet that can pursue its enemies for infinity—perhaps you’re forgetting one little thing: Your home, your loved ones, and your soft, delicious flesh are all now well within the range of battle-ready flying robots armed to the teeth and named after Satan.

Self-preservation instincts and infinite power supplies won’t help our robot adversaries, however, if they can’t reason at some level approaching human, and that’s our chief advantage. Of course there’s a substantial amount of research into artificial intelligence these days, but it’s all strictly ethereal—it’s not like that stuff’s got a body. There are chat bots and stock predictors and game simulators and chess-playing noncorporeal nancy boys in the robot kingdom, but even if a robot can crash the stock market, at least it can’t crash a car into your living room. Nobody’s stupid enough to give a rival intelligence an unstoppable robot body… right?

Uh… please?

Things That Are No Longer “Cute” When They Are Fortified with Steel and Enhanced with Crushing Strength

• Bumblebees

• Kittens

• Infants

No such luck. It turns out there are brilliant scientists hard at work doing exactly that: In 2009, a robot named the iCub made its debut at Manchester University in the United Kingdom and, much to the horror of mothers everywhere, it has the intelligence, learning ability, and movement capabilities of a three-year-old human child.

Does nobody remember “the terrible twos”? You know, that colloquialism referring to the ages of two to four, the ages when human children first become mobile, sentient, and unceasing little fleshy whirlwinds of destruction and pain? Well, now there’s a robot that does that, except it’s made out of steel and it will never grow out of it. The iCub can crawl, walk, articulate, recognize, and utilize objects like an infant. As anybody who owns nice things can attest, there is no exception to this rule: Infants can only recognize how to utilize and manipulate objects for the purposes of destruction. How long before military forces around the world attempt to harness the awesome destructive capability of an infant by strapping rocket launchers onto the things and unleashing them on rival battlefields to “play soldier”?

The iCub is being developed by an Italian group called the RobotCub Consortium, an elite team of engineers spanning multiple universities, who presumably share both a love of robotics and a hatred for humanity so intense that every waking moment is spent pursuing its destruction. And before you go thinking that the rigid programming written by the sterling professionals at the RobotCub Consortium will surely limit the iCub’s field of terror, you should know that the best part of this robot is that it’s open source! As John Gray, a professor of the Control