Imagine a world in which people coexist with systems that, just like us, have intelligence, thoughts,
Brave New World raises ethical questionspuzzles. How to behave with robots? Do we have a moral duty to those we have created? Maybe progress should be stopped altogether before artificial intelligence surpasses human intelligence? For now, such questions remain only speculation on the topic, because humanity does not yet have real examples of intelligent robots.
The film "Blade Runner 2049" tells about the near future, where the issue of attitudes towards androids has already been resolved. But not everyone agrees with this.
What other people experience remainsan unproven riddle for the observer: no one can be sure that his interlocutor really thinks exactly as the meditator himself thinks. And yet, it is customary to believe in the rationality of other people unconditionally. It is much more difficult to assume that robots have deep emotional intelligence.
How to believe that it is human?
To understand whether a particular AI can be mistaken forperson, he must be tested using the Turing test. It was created by computer pioneer Alan Turing in 1950. The question was not about actually replacing a person with a computer: the scientist was interested in the question of whether a machine could have visible intelligence.
In Turing's edition, the test looks like a dialoguebetween interlocutors, and its purpose is only to answer a single question: "Is your interlocutor a human?" One expert communicates with the computer. In modern competitions, there is a whole team of judges on the other side of the screen. They determine whether they are speaking to a living human being solely by their answers to questions asked. In addition, the judges communicate in parallel with real people. It turns out a kind of game of thimbles, in which the player needs to guess where the real person is “hiding”.
Until recently, chat bots were unable todeceive the judges. That changed when the Russian program, "boy" Eugene, finally passed the test. This happened a decade later than Turing expected - instead of 50 years, computers took 64 years. I must say that not a single car could pass the test one hundred percent. Moreover, it takes even less than 50% of the judges' votes to win - only 30%. Eugene seemed like a man of 33% of the experts. His conversation with people was short, only five minutes for each person. The rules, as you can see, are far from strict.
Communication style of Eugene Goostmanfar from human. The chatbot pretends to be a 13-year-old boy from Odessa, who is somehow able to answer questions about his personality, hobbies and knowledge. He is not as smart as he could be - excessive "well-read" AI only gives rise to doubts in experts. He also constantly asks questions himself, helping to develop the dialogue.
Not too strict criteria for entering the honorarythe class of "people" did not suit the scientist Hugh Lebner, who proposed a new "gold standard" Turing test in 1990. The new rules require that a person and a computer talk for at least 25 minutes, while the chatbot must have a long conversation with each of the four judges. The machine wins only if it can deceive at least half of the experts. Eugene couldn't pass that test.
There are other, more complex test variations.Turing. For example, an AI can communicate with three judges for two hours each. He has to convince two out of three experts. Futurist Ray Kurzweil and Lotus software company founder Mitch Kapor even made a $ 20,000 bet: Kurzweil bet that at least one robot can pass such a tough Turing test by 2029. The rate is still in force.
Can a robot be a subject of legal relations?
Human-like chatbots are not yet human. Developers of such programs believe that "they write not so much algorithms as novels." Behavioral imitation does not show real mental activity, creativity, association, and all that is inherent in the human mind and reason. In the issue of the rights of robots, we are not yet talking about mutual responsibility between the subject and the state, as well as other participants in legal relations. The law governing attitudes towards robotic systems cannot be absolutely human, and therefore the need for completely new laws appears before lawyers.
Legal systems around the world are built onthe assumption that only people can have rights. A clarification is needed here. Philosophers of modern times developed a system of natural and positive rights - inalienable from nature and created through a social contract. Only one creature on the planet gives rights - people. But those who can still be endowed are much more.
Today's Robot Rights Debateare summarized in a book by Northern Illinois professor David Gankel. It is called that way: Robot rights (from English "Rights of robots" - "High-tech"). The status of autonomous machines, he says, is constantly changing, because robots appear to be a "gray area" between individuals and objects. In fact, we already live in a world that is flooded with artificial entities with the rights and responsibilities of a “person”: this is IBM, and Amazon, and McDonald’s. They definitely don't become human, but somehow interact with the human race.
Movies like Terminator and 2001: A Space Odyssey ”convince viewers that robots can harm people. Of course, humanity is also capable of giving back to machines, although there are practically no stories when people become aggressors against robots. However, the American Society for the Prevention of Robot Cruelty (ASPCR) took the opposite idea as its basis. The organization was founded in 1999 but is still ahead of its time. The organizers do not think they are doing anything strange: "The inability to recognize the rights of non-human AI and grant them to it is analogous to the inability of colonial Western cultures to recognize the humanity and rights of non-European peoples." There is also a note on the site that in the 1890s, when the first animal welfare society was founded, its mission was also ridiculed.
ASPCR ceased to exist in 2016, but itsthe base, no matter how playful it is, has not disappeared. The European Parliament believes that the idea of robot rights deserves serious attention. A 2016 report by Luxembourg's Madi Delvo elaborates on the creation of a new legal status for an electronic identity to denote AI and robots. To be an electronic person means to have responsibilities and rights.
Let's imagine that a robot deceived a person and stolehis money in a transaction. The report offers an answer to the question of who should be held responsible in such a case. The proposed solution is to give the robots legal responsibility. At the same time, the creators or owners of robots must pay insurance premiums to government funds so that electronic individuals have the opportunity to cover costs.
The plan is based on the three laws of robotics,which was set forth by the science fiction writer Isaac Asimov. First, a robot cannot harm a person. Secondly, the robot must obey the orders of the person, if this does not lead to harm to other people. Third, the robot must defend its own existence as long as it does not harm the person. Finally, Azimov added "zero", the fourth law: a robot cannot harm humanity or do nothing in such a way that humanity will be harmed as a result.
Now the need to develop the rights of robotsseems far-fetched, but MEPs write that future possible autonomy raises legal questions now. “Ultimately, there is a possibility that within a few decades, AI will be able to surpass human intelligence and, if the system is not prepared, jeopardize the human ability to control its own creation - and therefore the ability to be in charge of its own destiny and the survival of the species,” - says the report. Therefore, even premature measures provide a field for thought.
In 2017, another EC report appeared, inwhich, among other things, was advised by the Robotics Civil Law Commission. It complements the same concept of electronic persons who can be liable for damage. It is assumed that robots of the corresponding status can make independent decisions or interact with third parties, except for their creators, independently.
Creation of a new subject of legal relations,more unusual, could be a loophole for manufacturers looking to evade liability for damage from robots. But robots do take on more and more everyday tasks, some of which are dangerous to others. And it's not about special assignments. For example, a drone might carry a box and accidentally drop it on a person's head. In human law, there is also responsibility for inaction: can AI be responsible for it?
What's happening with e-law in Russia?
The Russian economy lags behind in robotization, buton the other hand, Russian law takes measures in advance. So, at the end of 2016, Dmitry Grishin, a Russian entrepreneur and founder of Grishin Robotics, created the concept of the first full-fledged separate law on robotics. He co-authored with IT lawyer Viktor Naumov.
A robot is a device capable of acting, determining its actions and assessing their consequences based on information coming from the external environment, without full control from the side of a person.
Article 3 of the law of Dmitry Grishin
Grishin's concept proposes to equate the robotimmediately to animals and legal entities. Although autonomous machines do not have feelings, they can act independently, which means they fall within the scope of the concept of "responsibility." However, the law does not impute truly human responsibility to robots, but it says that for them it is necessary to create a separate register, similar to the register of legal entities. Robots can have manufacturers and owners. If the robot is initially potentially dangerous, then the responsibility for its actions falls on the "owner", and if a robot vacuum cleaner causes injury to a person, then the manufacturer will have to pay.
In addition, robots can become subjects of criminal law, but only in four cases:
- The robot was created to commit illegal actions.
- The robot was reprogrammed to harm a person by turning off its unit, which is responsible for safety when communicating with a person.
- The robot was originally created without this block.
- The robot was designed without realizing that it could be used to harm humans.
Dmitry Grishin's concept is based on fourthe laws of robotics by Isaac Asimov, which, according to the entrepreneur, have not yet lost their relevance. Also in an interview, the author of the law gives examples of robots for which the text will be important in the first place: unmanned vehicles and drones.
In addition to Grishin, robotic law in Russia is developingproject "# PravoRobotov". Also created at the end of 2016, the project is aimed at improving laws in the field of AI, data circulation and cyber security. The team deals with a wide range of challenges that the digital economy will have to face once it is built.
Not all Russian entrepreneurs are in the moodoptimistic. Alpha Robotic Venture CEO Vladimir Bely believes that it is pointless to create separate laws for the machine. In fact, they were and remain instruments, and the one who controls it is always responsible for the action of the instrument.
Nevertheless, the official authorities are holding the courseto develop a full-fledged bill on robotics. State Duma Speaker Vyacheslav Volodin said that a similar text would appear around 2022. The Russian authorities did not say anything more concrete about robotics, and it is especially difficult to predict something in 2020.
See also:
Comet NEOWISE is visible in Russia. Where to see her, where to look and how to take a photo
It turned out that made the Mayan civilization leave their cities
On day 3 of illness, most COVID-19 patients lose their sense of smell and often suffer from a runny nose