Robots are rapidly evolving - the time is approaching when they will fill the streets, offices, apartments and even our bodies. Are we ready to exist side by side with machines that behave like humans and have already begun to compete with us for jobs? Many humanitarian issues arise.
What can you entrust to a robot, and what is it better to refuse? Is he able to make decisions, feel emotions, does he have a certain free will and is he responsible for his actions? How to integrate robots into society? These questions seem naive. While. But it may turn out that the answers to them will decide the fate of our civilization.
When the lights went out, the hot melody that usually accompanies fashion models on the catwalk began playing. And we (the staff of Schrödinger's Cat magazine) saw two huge red heart-shaped eyes. The robot approached the audience with dignity and grace. The eyes turned into two blue circles, radiating good intentions and childish naivety, hands stretched out to people. The light was turned on again.
- Alantim! - the head of the Promobot project and the head of the robotics department of the Moscow Institute of Technology (MTI) Alexey Yuzhakov presented it to the public. - In 2012, we called him Promobot, then the offensive nickname Monument stuck to him, then he had the sonorous name Bogdan, but now he is Alantim - in honor of the creators of the new world Alan Turing and Tim Berners-Lee.
Photo: Skoltech / kot.sh
Alantim still turned back and forth and went to talk with journalists, showering them with jokes and stories about how robots will seize power on Earth and rename it Promobotia. A queue of girls immediately lined up to him, wishing to be photographed with such a cute creature. Alantim met the girls, answered questions about the meaning of life and the dollar exchange rate, and remembered everyone.
This is how the first working day of the new deputy head of the Department of Robotics of the Moscow Institute of Technology, the robot Alantim, passed.
Promotional video:
A car with character
Alantim is not a unique creature. The team from Perm, which created promobots, has already sold 60 such devices, and this year signed contracts for another 124, practically fulfilling the annual all-Russian plan for the supply of service robotics - in 2014, only about 200 domestic service robots were sold.
A promobot is an autonomous robot with character, designed to work in crowded places: promoter, guide, showman. He knows how to recognize some of the emotions of the interlocutor, is able to understand the question addressed to him and find the answer on the Web, and if he does not understand the question, he tries to laugh it off.
In the meantime, we went to the head of Alantim - MIT Vice-Rector Evgeny Pluzhnik
[Schrödinger's cat] How did you come up with the idea to hire a robot for the organization?
[Grigory Bubnov] We just realized that sooner or later the moment will come when robots will start hiring. And we decided to become pioneers.
[KSh] So it's not for nothing that they say that robots are claiming our jobs
[GB] This is actually not a joke, but a very serious threat, especially in areas where large numbers of people are employed. I think, for example, already in this decade autonomous cars will be on sale everywhere, and after that many drivers will be left without work. Governments will inevitably face this challenge - they need to start thinking about how to create new jobs and retraining programs.
[KSh] How do you think to use Alantim? Just like a showy toy?
[GB] Not only! For example, he has a research task - to accumulate data on the communication of robots with people, to study individual reactions to robots. This information will help to make the appearance and behavior of future modifications of Alantim more human and comfortable for communication.
Photo: Skoltech / kot.sh
[CS] What kind of information is he collecting?
[GB] First of all, data on reactions to his behavior - he has a system for recognizing emotions. For example, how do others react when he raises his hand? What if his manner of doing something scares them? Our gestures have been honed by evolution - we do not hesitate, let's say, scratching in the back of our heads - and with a robot, we still have to go all this way to "natural" behavior. By the way, he remembers people by photographing them.
But there can be problems: images of people are considered personal data, and there is a law restricting their use. In general, many questions of a legal nature will arise here - for example, is it necessary to pay income tax on the electricity that was given to the robot employee for charging? I think that soon we will see how the legislation in different countries will begin to change - along with the attitude of people towards robots.
Rubber Pushkin and other robots
The global robot market is already booming, sales are breaking records - tens of billions of dollars at the end of last year.
0.17% - This share of the world market for industrial robots is in Russia (according to 2013 data).
Photo: Skoltech / kot.sh
From heroes of Hollywood blockbusters, robots are about to turn into real characters - and we will begin to live with them. Will we get along? Starting with the horror stories about Golem and Frankenstein or the play by Karel Čapek, where the word “robot” first appears, all these stories, as a rule, ended with one thing - the uprising of man-made creatures against their creators. Can we create a reality that surpasses our fantasies?
An advertisement on the Promobot project website clearly explains why a robot employee is better than a human: does not need a salary and overalls, works without breaks, does not get sick, does not protest, does not get confused, does not say too much, does not need training … There are so many advantages, that, according to various estimates, in the coming decades robots can take from 40 to 80% of jobs, that is, we are threatened by an unprecedented wave of mass unemployment.
Perhaps, in the end, this will turn into a golden age, when, as Karl Marx dreamed, people will engage in free creativity. But we still have to live to see it - and preferably without social catastrophes. It is unlikely that it will be possible to significantly restrict the use of robots: never before has legislative bans been able to stop progress. What's the point? Rather, people will be banned from driving when robotic drivers become ubiquitous. Man is too unpredictable. Are not human pilots, for example, who are capable of directing an aircraft into the ground or into residential areas, dangerous?
That means we need to learn to live with robots
To begin with, we went to the place where there are many robots - to the international conference Skolkovo Robotics. A motley company of robots and their creators has gathered in the Skolkovo Hypercube. Big and small, flying and walking drones, professional firefighters and electronic wipers, mechanical arms and legs, torso and heads …
Here is Pushkin's rubber head, mounted on a tripod, as if on a stake, watching the passers-by and nervously twitching an eyebrow - there are too many people, eyes run up. The poet's face is so similar to the real one that you feel awe.
Photo: Skoltech / kot.sh
Nearby there is a robot that does not look like a human at all. Instead of a body, there is a stick on wheels, instead of a head, a screen that turns in all directions. This is a telepresence robot Webot, created in Skolkovo to help children with disabilities. You can connect to it via the Internet, navigate the building with it, see and hear what is happening around and interact with others.
Such robots are already used in several schools - children connect to them from home, and with their help attend lessons, complete assignments, answer teacher's questions, drive up to the blackboard, and communicate with peers.
The main consumers of service robots in Russia and throughout the world are people with disabilities. Robotics gives them hope for a return to full life - like the first Russian exoskeleton ExoAtlet. Its pilot Yaroslav, recently forced to lead the life of a paralyzed person, walks among the public like a cyborg from the future.
Another robot assistant was introduced to us by the Japanese professor Kazuhiko Terashima:
- Japanese society is aging rapidly, there are not enough staff in clinics. And the robot Terapio developed by us proved to be an excellent assistant to the doctor. He follows the doctor, carries the instruments, but most importantly, he remembers the data, which makes medical records unnecessary. In addition, he independently makes rounds of patients and in which case raises the alarm.
He has a childish expression on his face and a gentle voice, he knows how to laugh and cry, his big eyes blink all the time - in order to please patients, to cause them a reaction similar to that caused by children or kittens. Our robot is an important step in a robotic hospital project that continuously monitors patients. By the way, you know, now the operations are done with the help of robotic surgeons.
Photo: Skoltech / kot.sh
You can recognize an android by walking
Robotics and designers are performing in the conference hall of the Hypercube. Mikhail Lebedev of Duke University explains how he implants chips into the brains of monkeys that allow them to control robots with their minds. The session is moderated by Dmitry Teteryukov, professor at the Skolkovo Institute of Science and Technology, head of the space robots laboratory, who has worked in Japan for many years.
We, of course, peppered him with questions.
[Schrödinger's Cat] Statistically, the number of robots is growing rapidly, but where are they hiding - other than those used in industry?
[Dmitry Teterukov] In addition to industrial robots, an army of service robots is growing - everyone has heard of robotic vacuum cleaners or robotic quadcopters that deliver purchases. Many of these machines are used in medicine, and in the US Army they have been at war for a long time: for example, tracked robots with a manipulator are engaged in mine clearance. In general, they were very useful in emergency situations, they were used in the accident at the Fukushima.
[KSh] What kind of robots do you produce in Russia?
[JT] I'm continuing what I was doing in Japan. One of the projects is managing a group of robots, when they act according to a common plan, as a single flock, exchange information and take into account each other's actions. We, of course, will test all this here on Earth, but we plan that they will work in space.
[KSh] Science fiction writers have always described robots as such solitary creatures …
[JT] Distributing work among many is much more efficient than assigning everything to one. Let's say one self-driving car must process all the information about the road itself, but if such devices are everywhere, it is much more convenient for them to act on the basis of the general picture and program.
[KSh] When robots become human-like both externally and mentally, it starts to scare you a little?
[JT] There is a Japanese professor Hiroshi Ishiguro who creates androids that are almost indistinguishable from us - I recently mistook one of his geminoids for a human. But this is only while he is sitting. There has not yet appeared such a robot that can walk like a person - it is easy to recognize it by its gait. Ishiguro made both his own and his daughter's.
However, seeing her, the daughter was frightened, cried and ran away. That is, when a robot becomes very similar to a person, its rejection occurs. And when this threshold is passed, the unnatural appearance and movements are eliminated, the robot begins to be perceived as a real person. But to pass it, you need to achieve not only physical similarity, but also intellectual - the ability to conduct a reasonable conversation.
[KS] Is the famous Turing criterion, according to which a robot should be considered reasonable, which we cannot distinguish from a human in a conversation? Or does it need to be reformulated somehow?
[JT] Now this criterion is understood as a test for programs that simulate human chatting. But the point is not only in the intellect - in order to mistake an anthropomorphic robot for a human, it must be emotional, move smoothly, behave like a human being. There are already programs that very convincingly demonstrate sarcasm and irony. I think there will soon be separate tests for different abilities: the emotional Turing test will check the adequacy of emotions, and, for example, the motor test - the ability to move like a person.
Photo: Skoltech / kot.sh
[KSh] Aren't you afraid, roughly speaking, of a robot uprising?
[DT] After all, people write programs for robots, which means that they determine what robots can and what not, what desires they will have and what they will not. And why would robots go against their creators, for what reason?
[KSh] For example, because of our misunderstanding that they also need rights, or because they are taken apart and disposed of after their expiration date. And programs for robots will not only be written by humans - sooner or later machines will become self-learning
[JT] They will be trained within the framework we have set. It seems to me that the version of the future with the confrontation between humans and robots leads to degradation, no matter who wins in such a confrontation. An evolutionarily productive scenario is symbiosis, not a war of annihilation, parasitism, or other submission. Moreover, this symbiosis will also occur in the human body, so that the most interesting ethical issues, in my opinion, will arise in connection with cyborgization.
Where will the line between man and machine go when we start to install artificial organs for ourselves, install chips in the brain, improve our memory and other abilities with their help? It seems to me that a future awaits us without gadgets and personal computers - it will be enough to make a mental request or give a command, because the brain will be connected to the Network.
But maybe cyborgs will need restrictive laws, similar to the laws of Asimov's robotics - I even somehow formulated them. Their essence is that the machine in the cyborg does not seize power over the person, so that the decision-making power remains with the person, and not with his main processor.
[KSh] Do you have a dream robot?
[DT] This is probably an operator-controlled robot with a telepresence effect, which would allow full communication at a distance and convey to the operator all the sensations, creating the effect of complete presence.
[CS] Avatar?
[JT] Yes. I dealt with this topic before Avatar came out, but, of course, such films in some sense determine the future - our understanding of what we strive for and what is possible.
Photo: Skoltech / kot.sh
Moral machines
Meanwhile, in the Hypercube, a discussion began on the topic "Moral portrait of an anthropomorphic machine." Professor of psychology Maxim Kiselev gave an introductory speech on the moral character of robots.
“Roboethics deals with all humanitarian problems related to robots,” he explained. - For example, legal: who is responsible if a car without a driver drove somewhere - the robot itself, the owner of the car or a programmer? Or social, the main of which is potential unemployment. I myself have been to robotic factories more than once - for example, at a plant in Swindon, where components for MINI Cooper cars are made. It employs about two hundred robots and the same number of people, and before there were four thousand jobs. There are also specific moral issues, such as the so-called assistant robot dilemma. How do you know when machine assistance becomes counterproductive? Imagine a robot, soft and fluffy, following the child and collecting toys. But for the normal development of a child, it is necessary that he do it himself.
Filled with robophobia, we went to the creator of robotics, Professor Gianmarco Verugio, chairman of the science committee of the School of Robotics of the Italian Institute of Electronics, Information Technology and Telecommunications - he was just sitting nearby.
[Schrödinger's Cat] How did roboethics come about?
[Gianmarco Verugio] Once I was a fan of Asimov, by the age of sixteen I had read all of his books. Maybe that's why many years later something like an epiphany came over me, from which roboethics arose. I realized that the laws of Asimov's robotics - "A robot cannot harm a person" and so on - are not laws at all, but simply good wishes, none of them is observed in practice. The real law is e = mc2. Like other laws of nature, it cannot be broken.
Violation of legal laws is punished. And the three laws of robotics are like the biblical commandments that say "Thou shalt not kill": everyone seems to be in favor, but do not observe. Right now, robotic weapon systems are being developed in many countries around the world. Weapons are needed to kill, which means that these robots are created to kill people.
We must make the laws of robotics legally binding. Writing such laws is a difficult task. Suppose a robot can kill a person on purpose, or maybe accidentally, due to a breakdown or program flaws, it will often be difficult to distinguish between these situations. But it is worth starting to think about these problems now: if we go the wrong way, we can greatly spoil the future of humanity. It is foolish to create the future with closed eyes, but so far we are developing robotics blindly.
War robots are one of the main problems in roboethics. We are trying to draw the attention of various international organizations to it. This is not only my initiative, there is an international movement "Stop the killer robots!", Which campaigns on the Internet and is supported by the UN.
Another group of problems is related to the topic of biorobots and artificial organs. To what extent can we, for example, interfere with the work of the human brain? A friend of mine is experimenting with rats - inserting electrons into their brains and controlling their behavior. He's a real rat pilot! But when the human brain is full of all kinds of neuroimplants that enhance abilities or control cybernetic organs, then a similar interception of control can be done with people.
Photo: Skoltech / kot.sh
[KS] But we can't give up the benefits of technology just because new opportunities always go hand in hand with new dangers?
[DV] How to solve these problems so that technologies do not make us controlled objects of management, but, on the contrary, increase our freedom? I have no answer to this question, I just want to pose it, understanding all its complexity and incompleteness of our vision of the situation.
Many unexpected questions arise before roboethics. Imagine, for example, the following situation: a robot nurse is caring for a lonely elderly person who communicates with the car more than anyone else. And she quickly begins to consider her a person, love her as a person and suffer if she breaks down - we tend to do this even with primitive Tamagotchi, but here there will be much stronger feelings. Can such a robot be turned off? For me, a person is a person, and a machine is a machine, but I'm afraid people will soon start making mistakes …
[KSh] How do you tell them apart? Chatbots are already passing the Turing test …
[DW] I would start with the fact that our psyche is not just contained in the body, as in a container. The psyche is built into the body - we become intelligent and conscious as we gain bodily experience. The concept of artificial intelligence, coined by computer scientists in the 1960s, assumed that the virtual world was sufficient for it to appear. Now this idea is jokingly called GOFAI (good old-fashioned artificial intelligence), and it became clear to the developers that, in order to gain a semblance of intelligence, the program will need a robot body and long experience of interacting with the real world. But, generally speaking, we cannot make an intelligent machine, if only because we still have a very poor understanding of how the brain works, what reason and consciousness are.
[KSh] So, robotics doesn't deal with robot rights? But what will you do if tomorrow your computer says, "I'm scared, don't turn me off?"
[DW] Well, no, the rights of robots will start to worry me when the last person on the planet has all the rights. There is still too much work to be done for people to start caring for cars. Roboethics is about people, we must do it, because we create our own future.
Published in the popular science magazine "Schrödinger's Cat" # 5 (07) for May 2015
