Robots have been a staple of science fiction for nearly a century. The word “Robota” meaning “forced labour” first appeared in the 1920 play R.U.R: Rosumovi Univerzální Roboti (Rossum’s Universal Robots), which was penned by Czech writer Karel Čapek.
Since then the idea of robots as servants to mankind has blossomed from wishful thinking into an occasionally terrifying reality. The future envisioned in the 1950s and 60s promised a day where consumers (and particularly housewives) could kick off their slippers and pinafores and enjoy a gin and tonic while the robotic maid simultaneously cooked dinner, washed the car and beat the kids (it was acceptable back then). Of course, the robo-prophets such as Asimov et al may have been a little bit off on the dates, but in the first decades of the 21st century it seems that science fiction is making good of its promise. The robots are here at last … and some of them are incredibly lethal. Is this a cause for concern? Some would say so.
According to the Guinness Book of Records, the first death by robot occurred on 25 January 1979. Robert Williams, an employee of the Ford Motor Company at its plant in Michigan, USA, was “struck in the head and killed by the arm of a one-ton production-line robot”.
Of course, there was no malice on the robot’s part. Williams “reportedly climbed into the storage rack” to retrieve some auto parts when the robot, which was merely following its programmed orders, decided to occupy the same place in space and time as the unfortunate worker, resulting in the inevitable mess that occurs when solid steel meets squishy flesh.
A horrific milestone death
Two years later, at the Kawasaki Heavy Industries plant in Tokyo, Japan, engineer Kenji Urada was working on a production line robot on July 4 1981 when the errant droid pushed him into a grinding machine, killing him instantly. Urada had failed to switch the robot off fully, resulting in the grisly mishap. Happy Independence Day, Mr Urada - your horrific death will be hailed throughout history as an important milestone in the future robotic uprising.
By 2005, the figure of robot-related workplace fatalities had risen to 77 in the UK alone, and this trend is growing, as you would expect in an increasingly industrialised society.
But this figure applies only to “dumb” robots in the workplace, that is, machines that follow a strict, repetitive set of instructions. Over the last decade we have seen a significant increase in an altogether different category - robots whose prime function is to maim, kill and disfigure. There has been a significant rise in deaths in the 21st century from UAVs, more commonly known as drones.
According to The Bureau of Investigative Journalism, a UK-based non-profit organization, 2,400 people have been killed in the five years since the inception of Obama’s drone programme.
While drones are also considered “dumb”, there are concerns that these machines may one day be gifted with the ability to make life or death decisions without human interference. Current systems are referred to as “pilot-in-the-loop”, meaning that a human must have the final decision before an attack is made. With the increasing sophistication of AI, embedded systems, sensors, computer vision and logic algorithms, it is conceivable that one day a UAV or other unmanned system could have the capacity to become not only hunter and spy, but jury and executioner at the same time.
Human rights laws conundrum
So is this just some singularity fanboy’s fantasy, or is there a real risk of this scenario evolving? The United Nations seems to think that it is a very real possibility. In May 2014, 87 different countries convened at the UN’s Palais de Nations in Geneva for the Convention on Certain Conventional Weapons (CCW) Meeting of Experts on Lethal Autonomous Weapons Systems (LAWS).
The UN and the Campaign To Stop Killer Robots (yes, it’s a real thing) have been voicing concerns that use of robotic assassins may be in contravention of human rights laws and that, specifically, death by machine may be somewhat lacking in dignity.
“Is it morally acceptable to delegate decisions about the use of lethal force to such systems?” asked UN Secretary General Ban Ki-moon. “If their use results in a war crime or serious human rights violation, who would be legally responsible? If responsibility cannot be determined as required by international law, is it legal or ethical to deploy such systems?”
The outcomes of the Geneva discussions will be discussed at the formal conference of the Convention on Certain Conventional Weapons in November 2014, where member states will discuss possible next steps on fully autonomous systems. The results will be made available to the public shortly after. Whatever the results are, it is good to know that the issue is indeed being taken seriously, at the highest of levels.
As a fan of robots and as an aeronautical engineer I do have a certain love of UAVs, but as a pacifist the uses of these systems worry me. But fortunately all robotic systems want to crush the living crap out of us weaker carbon-based systems.
So how are these systems going to benefit humanity? Thankfully, it is not all doom and gloom. Advances are being made that will have positive effects on society. These spin-offs are being seen most prominently in the fields of AI, medicine, entertainment and transportation.
Just two months ago, for example, it was announced that a chatbot had beaten the Turing test for artificial intelligence by convincing more than 33% of participating scientists that it was indeed a real boy. This particular chatbot was the digitised spirit of a 13-year-old Ukrainian boy called “Eugene Goostman” and its victory immediately split public and professional opinion.
Professor Kevin Warwick, who helps run the Turing test, points out that it is a valuable litmus test for determining how far we are from creating an AI that we should be worried about. We have already seen the effects of heavily armed UAVs on populations in the Middle East and Professor Warwick is particularly worried about giving these weapons more autonomy than they deserve. He is also concerned about chatbots’ ability to dupe humans.
“Of course, the test has implications for society today”, says Warwick, a man best known for implanting chips into his body so that doors would open on his approach. “Having a computer that can trick a human into thinking that someone, or even something, is a person we trust is a wake-up call to cybercrime. The Turing test is a vital tool for combating that threat.”
Haters were quick to denounce the Ukranian chatbot.
They said that the test was not fair, that the programmers had given “Eugene” a backstory. They had created a human interest element that could convince 33% of judges that his mangled English and lack of cultural awareness was not a result of a badly coded bot, but of his tough childhood growing up without books near the irradiated wastelands of Chernobyl.
Whether you believe that Goostman passed the test or not, Warwick is absolutely correct - we need to keep a close eye on these things.
Cybernetic Kool Aid
Incidentally, Goostman’s inventor Vladimir Veselov is now chief of software development for Amazon. Given the online shopping giant’s recent revelation that it wants to start using delivery drones, it is clear that Amazon is totally drinking the cybernetic Kool Aid. It is certainly one way of cutting back on striking workers.
The medical world has been quick to make use of robotics. American company da Vinci Surgery has invested a not insignificant amount of time and money into developing a cybernetic surgeon. Its system is a remotely operated machine resembling a cross between one of George Lucas’ medical droids and an Animatrix-style torture device - in short, a lot of mechanical arms with scalpels on the ends. da Vinci promises high precision surgery that can remove human operator-related errors, such as hand tremors and fatigue. For example, a qualified surgeon can be located on land while the da Vinci machine can be operating on an oil platform, or other isolated location. The surgeon uses joystick-like controllers to command the appendages and any tremors in the human operator are quashed en route to the actual hardware, resulting in small, smooth and efficient incisions - in theory at least. In reality, it seems that da Vinci may secretly be playing for the other team, with reports of botched operations just coming to light. So far, at least 70 patients have filed lawsuits after experiencing complications from the robotic surgeon. Apparently one in every 15 patients experiences some form of injury after going under the knife. In one reported case, a patient was undergoing colorectal surgery when the overenthusiastic machine refused to release its grip on the patient’s bowel. The whole system had to be shut down and restarted in order for the beast to release its icy metal grip.
OK, so maybe da Vinci has a few teething problems, but it doesn’t require a large leap of imagination to envision a time when a fully autonomous robotic surgeon could perform simple surgical techniques. Like most of these examples, it should be seen as an indication of the shape of things to come. Presumably it will be accepted by the medical community at large when they can figure out a way to release a patient’s vital organs without resorting to CTRL+ALT+DEL.
Most of the machines discussed so far do not resemble our traditional notion of what a robot should look like. They have far too many rotor blades, tools and other weird mechanical appendages to be considered anything like what Asimov envisioned.
Luckily for us, there are several companies working to push us kicking and screaming into the trough of uncanny valley. That is to say, they are making robots that look and move like human beings.
Capable of mimicry
One company of note is US-based Hanson Robotics, which has been pushing the boundaries in terms of building lifelike robots capable of mimicking human expression. One such creation is Diego-san, who is pictured at the top of this article. Diego-san is an occasionally creepy little boy android, which has been designed using machine-learning methods that have analysed face-to-face interactions between mothers and their infant children. With high-definition cameras mounted in his eyeballs and computer vision algorithms in his bionic brain, Diego-san can recognise people, gestures and facial expressions, and is capable of learning directly from humans by using custom AI modelled on actual infants.
According to Hanson Robotics, the facial expressions of the cute little droid are important in establishing relationships and communicating intuitively with actual humans. Other Hanson creations include a lifelike robotic bust of Albert Einstein and a creepy disembodied head of cyber-prophet Philip K Dick. If you want to be simultaneously freaked out and impressed, then I urge you to mosey on over to YouTube and check out the videos of Diego-san and the other Hanson Robotics creations.
At present, Hanson Robotics is designing robots and animatronics for theme park and other entertainment purposes, but it plans to implement these systems in a more socially conscious manner, by deploying the androids to work in hospitals and with special needs children, such as those on the autistic spectrum.
Remaining on the topic of humanoid robots, we now turn to Google and its recent acquisition of Boston Dynamics. The American company had previously been responsible for the design of an actual Terminator-style, bipedal killing machine known as Petman, and was also behind the world’s fastest running robot, dubbed Cheetah, which has been clocked at speeds of above 29 mph. Petman was designed to test wear on chemical protection clothing for soldiers, being able to mimic human movement fairly accurately. In addition, Petman can climb stairs, operate other machinery and do a variety of exercises such as lunges and, presumably, star jumps. The main contract holder before the Google acquisition was DARPA, the Defense Advanced Research Projects Agency, and post-acquisition Google has been distancing itself from military contracts.
If real life were a movie, then Boston Dynamics is the genuine Cyberdyne Systems, the fictional company that brought cybernetic hellfire to the population of Earth in the Terminator franchise. Although the reasons for the acquisition remain mysterious, it is perhaps comforting to know that Skynet is now out of the hands of the generals and safely in the hands of the nerds. Unless you happen to be a Google-phobe, in which case you should stock up on tinned food and get your ass into your nearest bunker, toot sweet. For the last year, Google has been on a bit of a spending spree in terms of robotic companies.
Whether you love Google or loathe it, it is pretty fair to say that any major developments in the field of consumer robotics over the next decade will be coming from the big G.
So, the future may not belong entirely to the robots, but we are sure as hell going to have to share some of our living space with them. It has been suggested that the first jobs that will be lost to a replacement cybernetic workforce will be those in the service industry. This video shows a tray-collecting robot at a KFC restaurant in Mainz, Germany. If these machines are given the gift of judgement, then you may be wise to tip graciously lest you feel their vengeful appendages gripping onto your stingy skulls!