by Marc Goodman
Importantly, the rise of the criminal UAV is completely incompatible with our current security paradigms. Prisons use tall, sharp, often electric fences to isolate criminals for reasons of public safety, a system that worked relatively well for hundreds of years. But our security and defense mechanisms were meant to protect us from offending human criminals, not robotic ones. It may be time to rethink that. Drones can circumvent not only prison fences but any fence, including those protecting your backyard, office building, or even national borders, as the narco-syndicates of Latin America are demonstrating. In Mexico, for example, Crime, Inc. has hired assembly-line workers from local aircraft factories who moonlight for the cartels designing UAVs. In the Santa Fe district of Mexico City, just near the Bombardier factory, a covert narco-drone factory was uncovered, according to Mexico’s Public Security Secretariat. Building on American, European, and Israeli designs, these ultralightweight autonomous aircraft are much larger than the average quadcopter, weighing a hundred pounds, and have foldable wings so they can easily be transported by and hidden in trucks on either side of the border. They fly low and are undetectable by radar. Each drone can carry a hundred kilos of cocaine per trip, coke that costs $1,700 a kilo in Colombia, $8,000 in Mexico, and $30,000 in the United States, netting traffickers more than $2 million a flight. Since 2012, the DEA has documented at least 150 such narco-drone crossings carrying multiple tons of cocaine. With profits like this, cartels from Cali to Sinaloa are reinvesting their proceeds into advanced research and development, spending millions to ensure a much more prominent role for their criminal robotic workforce moving ahead.
Apart from drugs, there are much more troubling items criminal techies can attach to drones, including firearms. YouTube is already replete with hobbyists showing self-built remote-controlled flying robots that perform elaborate tasks such as tracking and shooting people with water pistols and paint balls—a pastime ideal for criminal or terrorist adaptation. Other videos show hobbyists flying drones with stun guns on board, shooting electrical prongs at their prey and Tasing them down to the ground with eighty thousand volts of electricity. But the trend of course does not stop there; real guns have been used too. The earliest video of an actual firearm, a .45-caliber handgun, mounted on a remote-control helicopter and firing appeared way back in 2008. Since then, numerous other videos showing smart-phone-controlled UAVs with guns have appeared online, including an HD version of an octocopter carrying a .45 Colt handgun fired repeatedly using a remotely controlled robotic trigger finger. With so-called automatic “follow me” technologies, they can even autonomously track a particular individual as he runs down the street. Using a smart phone to shoot an actual gun mounted on a flying robot that costs a few hundred dollars means that first-person shooter games have entered three-dimensional space and become a reality. How long will it be before a criminal or mentally unstable individual uses such a device to kill others?
For as dangerous and frightening as such a scenario might be, other, much more pernicious payloads can be loaded on board UAVs as well, including explosives and even weapons of mass destruction, such as a biological, chemical, or radiological weapon. For under $20, bomb-drop systems for remote-control aircraft are available online, similar to the bomber doors on military aircraft that open up when directed via remote control or when reaching a particular GPS navigation point. Might drones become the next suicide bombers? Al-Qaeda, Lashkar-e-Taiba, and numerous other terrorist organizations already have active drone development programs. Several YouTube videos have shown farmers, tired of working in the sweltering heat, who have converted their remote-controlled helicopters into crop dusters. If a terrorist were to use the same concept to spread a lethal agent over a crowd instead of pesticides over rice fields, the potential for harm could be enormous.
Drones, as the military has shown us, can also be used in highly targeted ways against particular individuals, whether for personal revenge, criminal attack, or terrorism. We’re already starting to see high-profile people come under assault in both strange and dangerous ways. In late 2013, Chancellor Angela Merkel of Germany found herself under drone attack during a campaign rally in Dresden when a quadcopter UAV charged toward her onstage, only to crash at her feet. The attack was carried out by Germany’s Pirate Party, which said that it wanted to make sure the Chancellor knew “what it’s like to be subjugated to drone observation.” Her security protective detail surely got the message. Though nobody was injured as a result of the stunt, the story could have had a much unhappier ending had the device been armed or carrying explosives.
Drones can also cause damage when launched at other conveyances, shocking automobile drivers and causing crashes. There are already numerous reports of hobbyists intentionally flying drones directly into the flight path of jet aircraft around the world, causing pilots to take drastic evasive actions to avoid collision, including planes from American Airlines, US Airways, Alitalia, and Virgin Blue. Had any of these flying robots been sucked into the airliner’s engines, they could easily have caused a crash similar to the one that downed the US Airways jetliner into New York’s Hudson River. As time marches on and flying, swimming, rolling, and walking robots enter our lives, we will have to figure out ways to safely and peacefully coexist with them, but the future of robotics itself may bring with it even greater risks to be managed.
The Future of Robotics and Autonomous Machines
Robots will become faster, smarter, and smaller. Already great advances are being made in micro-robotics. The devices, some small enough to fit on your fingertip, are flown remotely and can be equipped with an HD camera and microphone, taking surveillance privacy concerns to a whole new level. Dragonfly drones were already reportedly used to spy on antiwar protesters in Washington, D.C., back in 2007, and the air force unveiled robo-bumblebees that could not be detected in hostile environments while flying into buildings to “photograph, record and even attack terrorists.”
Another forthcoming breakthrough in robotics will be that of “swarm” capabilities—getting multi-robot systems to act in unison via collective behavior that mimics nature in the same way ants cooperate and birds flock. Using advanced distributed computing power to self-organize and solve problems, swarms of bots could coordinate their efforts to achieve incredible things whether in disaster relief, search and rescue, oil spills, or manufacturing. Much progress is being made in swarm intelligence, and in mid-2014 researchers at Harvard University created the largest robot swarm ever, using 1,024 tiny robots the size of a penny that could find one another and collaborate to assemble themselves into various shapes and designs, including stars and letters, like a mechanical flash mob. But there may be a gathering swarm on the horizon when it comes to self-organized cooperating bots: they can also be a force for ill. One drone with a handgun chasing you down the street is bad enough, but a swarm of thirty of them is terrifying and likely unsurvivable. Moreover, once swarming robots come into common use, any hack they suffer or virus they get could be disastrous, as it would affect all the bots on the network, just like the scenes portrayed in the television show Star Trek wherein the crew of the USS Enterprise uses a computer virus to successfully destroy the Borg collective of cybernetic organisms, except we might be the Borg that is destroyed. When the military starts widely using armed UAVs operating in swarm formation to attack the enemy and they get a virus (as the U.S. drone command already has), how easy might it be to turn the weaponized flying robots against their masters or innocent civilian populations?
Moving ahead, not only will we have ever-smaller bots swarming about us, but we will have robots that are increasingly autonomous—intelligent machines capable of performing tasks and making decisions in the real world on their own without explicit human control. An autonomous robot, like the Roomba vacuum, makes decisions based on its programming, but on its own in real time, using “bump and go” algorithms to move and avoid obstacles, enabling it to analyze and adapt to unfamiliar environments.
It is, h
owever, with military robotics that particularly difficult questions about autonomy arise: How much is too much? A ground-based robotic army medic that can autonomously rescue a wounded solder from the battlefield and provide lifesaving first aid sounds like a great idea. But a UAV that can find its target and autonomously make the firing decision to kill may give many pause. Yet that is exactly where we are heading. As robotics, artificial intelligence, and computer processing speeds improve exponentially, at some point human beings just aren’t going to be fast enough to keep up, particularly in the realm of warfare. Once your enemy goes to fully autonomous warfare, you will be compelled to do the same or face destruction.
Though they may seem like the stuff of apocalyptic dystopian films like The Terminator, autonomous killing machines are already here. The BAE Systems Taranis is a fully autonomous aircraft that can “fly deep into enemy territory to collect intelligence, drop bombs and ‘defend itself against manned and other unmanned enemy aircraft.’ ” In the demilitarized zone across the Korea Peninsula, South Korea has deployed Samsung SGR-1 border control sniper robots that can detect intruders with heat and motion sensors and automatically fire on targets up to a kilometer away with their embedded 5.5 mm machine guns and 4 mm grenade launchers. Though the border bots currently require human permission to attack as a matter of policy, technically they can go fully autonomous at the flip of a switch. Lethally autonomous killer robots will take many forms: walking, swimming, flying, and driving machines that can chase their prey, or just lie in wait. Yet despite our growing technical capability to outsource kill decisions to machines, doing so is fraught with a panoply of legal, ethical, moral, technical, and security implications.
Industrial accidents with robots are bad, but as we saw with the horrific South African National Defence Force computer glitch, accidents involving robots with automatic weapons can be catastrophic. As robots proliferate, we will suffer the consequences as Moore’s law clashes with Murphy’s law. Poor programming, inaccurate data, and software errors will undoubtedly lead to tragedy when robots themselves can decide to kill. Moreover, armed robots connected to the IoT will be hackable, as will their safety features and protocols, adding yet another notable peril to be considered. So too is the matter of repressive governments using killer robots to quell dissidents or narco-cartels to kill cops and rival drug gangs. It may seem like hyperbole to suggest that Crime, Inc. will one day have autonomous killing robots, but of course it will, just as it has adopted any number of other prior military technologies, including night-vision goggles, the Internet, and UAVs. Experts in both human rights and technology are concerned about delegating kill decisions to machines. The topic has also been raised by the United Nations, Human Rights Watch, and new organizations such as the International Committee for Robot Arms Control and the Campaign to Stop Killer Robots. The science fiction author Daniel Suarez and the roboticist Noel Sharkey have both even given impassioned presentations on the topic at TED calling for a global prohibition on robots’ autonomously killing or injuring a human being, a rather sensible idea indeed, as Asimov first proposed decades ago.
That robots will enter our lives for everything from elder care to meal preparation and surgery is a given. They may well be a tremendous force for good, but as we have seen throughout this chapter, robots are also being used by street thugs, Peeping Toms, narco-cartels, and terrorists, a trend that will surely accelerate as their functionality improves and their prices drop, particularly in response to incredible new and complementary technologies such as 3-D printing.
Printing Crime: When Gutenberg Meets Gotti
Restrictions are difficult to enforce in a world where anybody can make anything.
HOD LIPSON
3-D printing, or, as it is sometimes called, additive manufacturing, promises to bring the Star Trek replicator to life. At the push of a button, a magical machine can make physical objects before your very eyes using a wide array of materials, including plastic, metal, wood, concrete, ceramics, and even chocolate. Just as you can send a photograph to your 2-D ink-jet printer, so too can you download or create a design on your laptop and send it to a 3-D printer, which, using a variety of techniques, can build objects in three dimensions, layer by layer, with incredible precision. These digital manufacturing techniques are making it easier and cheaper to build not only robots but a whole host of products ranging from airplane parts to fully functioning SLR cameras and lenses.
Goldman Sachs has noted that when compared with traditional manufacturing, 3-D printing will drive greater customization and reduce costs for complex designs, and others have forecast a 500 percent growth in the 3-D printer market to $16 billion by 2018. Today, inventors such as Scott Summit, founder of Bespoke Innovations, are using 3-D printers to create next-generation customized prosthetics that are not only perfectly fitted but beautifully designed. Digital fabrication can be used to print entire homes, concrete, electrical wiring, plumbing, and all. NASA has even purchased a 3-D printer for the International Space Station from the Silicon Valley start-up Made in Space to ensure it never has to worry about a missing part on board endangering the lives of astronauts, as was the case with Apollo 13. Bio-fabricating printers have taken things to the next level with machines that can even print human tissues and organs, such as capillaries, kidneys, ears, and hearts, potentially doing away with organ transplant lists and saving lives.
Prices on home 3-D printers—machines that used to cost tens of thousands of dollars—are dropping precipitously, and models such as the popular Cube 3 made by 3D Systems can be purchased at Staples for $999. Amazon has created its own 3-D printing store, and Web sites such as Thingiverse have become the go-to locations for users to freely share and customize their digital design files to make everything from jewelry to iPhone cases, and MakerBot offers kits to even build your own 3-D printer. Autodesk’s free 123D software and apps can turn any 3-D digital model into a real-world object, and its open-source Spark operating system may do for 3-D printers what Android did for smart phones. These developments may shift manufacturing away from mass production toward mass customization, where people can print the exact shoe, table, or toy just as they like it. Chris Anderson, the former Wired editor in chief, has deftly documented this so-called DIY maker movement in his book Makers, in which he cites open-source design and digital manufacturing as the foundations of the new industrial revolution.
Another remarkable aspect of 3-D printers is that the devices are moving toward total self-replication. Today most 3-D printers can print more than 50 percent of the parts required to make another 3-D printer—a percentage that is increasing rapidly. The printers allow physical objects to be transmitted over the Internet and printed on demand. 3-D printers, just like robotics and the IoT, are ushering in the age when the analog and the digital are merging and becoming indistinguishable from each other. Bits and bytes are becoming atoms, and 3-D scanners such as Microsoft’s Kinect can turn physical objects back into ones and zeros. The result may well be massive disruptions in manufacturing, retail, and even geopolitics. Local fabrication and assembly could have deeply positive impacts for the environment. When you can print the things you need at home, why run down to the local store? And if American firms can print more of what they need here, does it make sense to have tons of cheap plastic crap shipped across the oceans from China? Regardless of how these transformations play out, there is one group of individuals who have already wholeheartedly embraced the maker movement: Crime, Inc.
Just as robotics has brought new cyber risks into our three-dimensional world, so too will digital manufacturing. The first area criminals will pursue in the world of 3-D printing is intellectual property theft. Previously, it was only digital intellectual property that could be perfectly pirated and duplicated—music, video, games, and software programs. That’s about to change. Though crooks have been making fake Gucci handbags and knockoff Cartier watches for some time, they were relatively easy to spot because of shoddy design and cheap manufactur
ing. But in the future these objects will easily be subjected to ultrahigh-resolution 3-D scanning and printing, making copies visually every bit as good as the original. The Gartner group has already predicted that 3-D printing will result in over $100 billion in intellectual property loss globally, per year, by 2018.
Digital manufacturing will also be a boon to burglars and stalkers who can now just take a high-resolution photograph of the home or office keys that you casually left on your desk and use a service like KeyMe to have duplicate keys printed via the 3-D printing marketplace Shapeways. There are apps too, such as Keys Duplicated, which will do the same thing, providing the keys to your castle to more people than you might like. If this bothers you, you’re not alone. In 2012, cops uncovered computer-aided design files online that allow criminals to digitally manufacture police handcuff keys, including ultra-secure models whose manufacturers do not sell keys to the public. In the future, your drug dealer may be a printer as well. Scientists have already developed a “chemputer” that can print medicines such as ibuprofen on demand. While the potential humanitarian benefits are enormous, it won’t be long before Crime, Inc. adapts these machines for meth, crack, and Oxycontin, vastly simplifying its supply chain and distribution issues.
Perhaps one of the greatest controversies surrounding 3-D printers is the ability to produce firearms, and perhaps no man has done more to make that a reality than Cody Wilson, a twenty-six-year-old former law student, anarchist, and libertarian à la Dread Pirate Roberts. Wilson created the Wiki Weapon Project, brought us Darkwallet and its untraceable crypto currency, and founded Defense Distributed—a nonprofit online designer, publisher, and repository of firearms blueprints that can be downloaded and printed with a 3-D printer. Among his 3-D printed creations was a lower receiver for an AR-15 semiautomatic rifle from which he successfully fired six hundred rounds of ammunition. The lower receiver is the key part of the firearm and the only one regulated by law; the rest of the parts in many states can be obtained without background checks or even identification. In May 2013, Wilson also designed the Liberator, the world’s first fully 3-D-printed gun, designed to fire standard .380 handgun bullets, and 100,000 people around the world downloaded the drawings. When asked by the press how he felt about his accomplishment, Wilson replied that now “anywhere there’s a computer and an Internet connection, there would be the promise of a gun.”
