Should a social robot be human-like? Social robots Who is a social robot

The picture of interaction between man and machine is becoming clearer and more interesting. Here you are lying with the flu for many days now, and all this time a caring robotic nurse or maid is giving you medicine in bed. She also becomes an excellent assistant in caring for children. Isn't it tempting?

This will be our future, as the development of robots that imitate human social properties and relationships is in full swing. At the same time, developers and experts describe the problems encountered along the way...

Tasks of social robots

The social category includes those robots that can understand humans and themselves in a social context. They are communicative, easy to train, and their behavior is as similar as possible to humans. In social robotics, the emphasis is on the integration of machines into social interactions, many developments have already been included in various areas of our lives, and there are more and more of them every day.

Social robots can be divided into several areas:

  • robotic home assistants;
  • robots included in the professional sphere (firefighters, astronauts, soldiers);
  • robot teachers and assistants.

Companion robots can be classified as a separate group. Their task is to provide friendly support to a person, causing positive emotions and having a therapeutic effect on his psyche.

Another separate group can include social robots, which are part of the media sphere and largely perform entertainment functions. We are talking about robot artists, singers, musicians, some of whom are already actively touring the world.

More recently, information has appeared about the release of a new, first family robot, which will become not just an assistant and friend, but a full-fledged member of the family. can react to a person’s mood and try to improve it: make faces, tell jokes or play music. This is a robot that can find an approach to each family member and meet their needs. He will report an incoming SMS, and will deliver it directly to the owner (he will not confuse messages for his daughter and father), and will help in the kitchen (he will find a recipe for the desired dish on the Internet).

Everything sounds wonderful in words, but before social robots enter our home, we must be sure that this process will not have negative consequences. The editors decided to figure out what obstacles might arise when introducing social robots into everyday life. After studying tons of information, we identified the following problems.

Problem: high cost

Before interacting with the robot, you must purchase it. And here we have to deal with the high cost - one of the reasons why robots have not yet appeared in every home.

As experts explain, the high cost is largely due to the high cost of each part. A robot is required to be “smart,” adaptable, etc. For example: a robotic arm. It can be the simplest, moving up and down, or it can have super sophisticated functionality, making movements along different trajectories. This increases the price of the technology several times. But it is possible that the pace of technology development will make it possible in the near future to reduce the cost of a social robot.

Problem: The Uncanny Valley Effect

Even if you have the financial opportunity to purchase such a robot, it is possible that you will immediately fall in love with it. - another problem that developers will have to find a way to solve, otherwise humanoid robots will remain rather dubious friends. Bright minds have repeatedly tried to create a robot that is as similar to a human as possible, but so far all attempts have failed.

The Japanese social robot CB2 was supposed to become as child-like as possible. SV2 has lively eyes and gray skin, but he never turned out to be cute.

And the Japanese KOBIAN with an expressive face, brow ridges and plump bright red lips is increasingly called a “monster” in the press.

One way to avoid this effect was found by the creators of the Nexi robot, which, by the way, is included in the list of “Best 50 Inventions” according to TIME magazine (2008). The success was achieved by transferring animation to real time, which Personal Robots Group specialists gleaned from the book “The Illusion of Life - Disney Animation”. Thus, Nexi contained the features of the so-called “exaggerated humanity”. If a robot wants to reach for an object, its hand does not jump forward mechanically and flawlessly. First, the robot focuses its gaze, then takes the desired pose and swings smoothly. This is very effective psychologically.

Many people have already managed to communicate with Nexy, including residents of nursing homes. They believe that through genuine communication and the inclusion of people in this process, society perceives their robot as a partner.

Experimenting with Nexy, we were convinced that if the robot’s behavior corresponds to generally accepted ideas about trust and friendliness, then others react to it in the same way as they would react to people. We hit the same buttons,” shares Cynthia Breazeale, head of Personal Robots Group.

Problem: emotional attachment

If you succeed in establishing trust, it is possible that you will have to face a new problem - emotional attachment to robots.

It seems that we are simply defenseless against emotional attachments to certain objects. So is it decent for us to speculate on this? - Sherry Turkle, director of the Technology and Personality Project.

Sherry Turkle's company conducted a study to find out how strongly people become attached to social robots. For these purposes they used the Paro robot.

Paro is a baby seal-like robot that was created for therapy purposes. He moves cutely in your arms, purrs when you pet him and becomes quiet when he connects to the network using a pacifier.

According to Sherry, children playing with Paro treat it like robotic dolls that are sentient and have emotions. But she was even more alarmed by the role of this machine in the lives of old people, who are most often lonely, and therefore become attached to the social robot even stronger and faster.

We give a robotic nanny to our children and the elderly, while we ourselves re-read the latest news on the Internet, comments Sherry Turkle.

Many researchers agree with her that the widespread use of social robots may pose some danger to the human psyche.

Problem: Unforeseen situations

But this is not as scary as the possible emergencies that you will have to face while living side by side with the robot. For example, such a development as a robot vacuum cleaner, which was introduced into our lives earlier than others, has more than once given cause for concern. There were reports in the media that an automated vacuum cleaner independently called the police, tried to escape from a department store, and even attacked the owner. Of course, these are all isolated incidents, however, they lead to various thoughts.

Here's a simple example: a patient with autism decides to attack a robotic caregiver and break it. What should be the machine's response? It may turn off, but in this case it threatens complete destruction for a very expensive device. The machine may begin to defend itself, but where is the guarantee that this will not lead to physical or moral harm to the patient? The developers still have to work on solving these problems.

We still dream of a robot that could help us and at the same time have sufficient charisma, - Francesco Ferro, CEO of PAL Robotics.

At this stage, social robots for the most part are present in single copies, communicating with people only at exhibitions or in laboratory conditions. Therefore, a lot of time will still pass before the future promised by science fiction writers, and developers will have the opportunity to cope with emerging problems.

Still from the movie “Blade Runner 2049”

On January 14, 2018, a discussion took place at the Poltava Discussion Club on the topic: Should a social robot be like a human?

Arguments for the position of the statement (must be similar) prepared and voiced Evolutionary(discussion nickname), arguments of the objection position (should not be similar) - Quixote(debatable nickname).

What is a social robot?

A social robot is a robot that interacts socially with people. For example, a robot consultant, an assistant, a robot nanny, a teacher, a nurse, a doctor, an actor, a singer, a dancer, etc.

Therefore, we asked the question: That is, in fact, should he be outwardly our human copy?

We solve this issue using the Discussion Method according to the Rules of the Poltava Discussion Club.

Video discussion

Arguments of the parties

Significantly shortened for this publication, full arguments on video:

First, blue argument “For”

Evolutionary: The similarity of a social robot to a person makes communication easier.

When it comes to social robots, ...their appearance matters a lot, since most people tend to trust people. People expect to see the emotions of their interlocutor in order to show their own. This is impossible if the social robot is not human-like. The humanized appearance of the robot contributes to its animation on the part of the person communicating with it; the person perceives it as one of their own.

First, blue argument “Against”

Quixote: The Uncanny Valley Effect.

The "Uncanny Valley" effect is the hypothesis that a robot or other object that looks or acts approximately like a human (but not exactly like a real one) causes hostility and disgust in human observers.

[We studied] people's emotional reactions to the appearance of robots. At first, the results were predictable: the more a robot resembles a person, the more attractive it seems - but only up to a certain limit. The most humanoid robots unexpectedly turned out to be unpleasant to people due to minor inconsistencies with reality, causing a feeling of discomfort and fear.

The reason for this psychological phenomenon has not yet been clarified. There are versions. [...]

The next explanation is the “psychopath theory.” She says that we fear not so much that we are incapable of empathy, but that the robot itself is incapable of empathy - in other words, we perceive such a creature as a psychopath.

Second, green argument “For”

Evolutionary: The development of humanoid robots advances medical technology and psychology.

In scientific and technological processes, value is often not only the goal itself that needs to be achieved, but also side discoveries and technologies. The process of creating a humanoid robot is not only programming, but also modeling. In this case, modeling of a person - movements - that means joints, facial expressions - modeling of the skin covering, facial muscles, temporomandibular joints. In fact, electronics and mechanics are tasks simpler than creating a social interface for a social robot. Here it is also necessary to study and model facial expressions and speech. And speech is not just a pleasant voice, it is its emotional coloring. [...]

By creating a model of a person in the guise of a social robot, scientists will make more than one useful discovery.

Second, green argument “Against”

Quixote: Flies separately, cutlets separately, or know your place.

So, let's look at Ernst Hoffmann's short story “The Sandman”. The hero of the novel, Nathaniel, falls in love with the mechanical doll Olympia, which ultimately leads him to madness. Naturally, Olympia always listened to Nathaniel so selflessly, nodded to him, constantly repeating “Ah, ah, ah.” It was so sweet, so natural that Nathaniel did not notice the catch. And not only our hero, but also other people did not notice that Olympia was just a doll. The ending of the novella is sad: Nathaniel jumps from the town hall tower and falls to his death.

And people become distrustful of each other in society. Therefore, many lovers, in order to make sure that they were not captivated by a wooden doll, demanded that their beloved sing slightly out of tune or dance out of tune.

To prevent such incidents from happening, it is necessary to clearly define that robots are not people, they are completely different creatures. Therefore, they should look completely different from people.

Third, red argument “For”

Evolutionary: Animation and gender perception are important for humans.

People tend to animate technology and endow it with human qualities. ...Robots that meet our expectations are perceived more joyfully and easier. And since a person is a creature divided into two sexes, a gender stereotype is also superimposed on his perception of the robot.

For example, people perceive a male robot more harmoniously in the form of a security guard or a foreman. And the female image is more suitable for robot nannies, cooks and secretaries. For example, the famous robot Flobi is presented in two different versions - male and female.

[...] That is, a robot that has the shape of a human body fits best into our world.

Third, red argument “Against”

Quixote: The shape of an object reflects the functions it performs and arose evolutionarily.

The form of both humans and animals was formed in the process of evolution, by the method of natural selection. That is, we see what survived the process of evolution, what turned out to be the most viable. We're used to it. In this regard, the form of a person or an animal is quite functional. All organs of the body are needed to perform any vital functions. The form contributes to this in the best way.

[...] If we take a social robot, then its purpose is to perform some social function. For example, the purpose of the interlocutor robot is to speak. This is its function. No others. Such a robot does not need a liver or kidneys, because it will not drink beer with friends.

And if it is intended for conversation, to perform this function it will need the so-called “organs” for vision, for listening, for producing sound. And these “organs” must be placed on some kind of base - a kind of head. [...]

Summary of the parties' arguments

Evolutionary: the summary of my arguments that a social robot must to look like a person:

  1. The similarity of a social robot to a human makes communication easier.
  2. The development of humanoid robots contributes to the development of medical technology and psychology.
  3. Animation and gender perception are important for humans.

Quixote: the summary of his arguments that a social robot should not to look like a person:

  1. The "uncanny valley" effect.
  2. Flies separately, cutlets separately, or know your place.
  3. The shape of an object reflects the functions it performs and arose evolutionarily.

Overall result of the discussion

Robots are not people, but completely different creatures. Hence the conclusion that they should look different. Moreover, the shape of the human body naturally reflects the functions it performs. This form is also economical, expedient, and has been perfected over millions of years of evolution. For a robot, it will be both uneconomical and impractical. After all, such a form will not reflect the essence of the robot.

But due to the pursuit of perfection, people try to make robots similar to themselves. We assume that this will lead to many useful inventions for our overall further development.

Vote

Alexander ZOLOTUKHIN, organizer of the Poltava Discussion Club

About 30 years ago, many popularizers of science, as well as authors of science fiction works, believed that in our time robots would actively help people. But the robotization of everyday life, so to speak, is not as active as it was thought.

True, robots are still gradually penetrating our lives. Smart microwaves, smart TVs, robotic vacuum cleaners, all this already exists and continues to develop. True, you can’t communicate with such devices, but you’d like to have a robot that you can talk to, right?

And such a robot appeared. We are talking about the robot Jibo, whose developers (the team comes from MIT, Massachusetts University of Technology) position it as the first social robot for the family.

What can a robot do?

Thanks to good technical equipment and appropriate software, Jibo is able, for example, to recognize all family members and find its own approach to each;

The robot is sensitive to a person’s mood and behaves accordingly, trying to support a person in a bad mood;

In addition to communication, Jibo can serve as an assistant, notifying you of an incoming message via email or SMS. The robot is equipped with wireless communication modules and can read information from any device connected to it (at the owner’s request). At the same time, it notifies the “right” people about messages. In other words, the robot will not confuse the SMS messages sent to the daughter and the father, and will not tell the father that “your cat is already waiting for you at the club,” for example;

The robot can also help in the kitchen, receiving information about a particular dish from the Internet (the robot also connects to the Internet, yes). Thus, instead of reading the recipe from the book, you can ask Jibo about it;

Well, Jibo entertains its owners quite well, demonstrating its affection, making funny faces and playing music. A robot can tell a child a bedtime story, for example, or tell a funny joke to parents.

Technical equipment

Jibo is quite well equipped; the developers took care of its hardware.

  • Case material: aluminum, ABS plastic, glass;
  • Display: LCD HD display;
  • Movement: the robot is capable of making movements along 3 axes, with rotation (we are talking about the robot’s head, it cannot move on its own);
  • Sensors: two color stereo cameras, circular sound localization, touch housing and display;
  • Sound: 2 high-quality speakers, advanced acoustic system;
  • Backlight: full spectrum LED;
  • Communication modules: WiFi and Bluetooth;
  • Processor: ARM processor;
  • English language;
  • Dimensions: height - 28 centimeters, width - 15 centimeters;
  • Weight: 3 kilograms.

Of course, the Jibo robot is not yet the robots that Asimov wrote about, with developed intelligence and a positronic brain. However, Jibo is able to communicate with the owner, the device establishes feedback with the person, and tries to adapt as accurately as possible to each person.

Price

Of course, the price of the robot cannot be low (taking into account at least its equipment and software). The developers valued their efforts at $499. You can order from

Robot nurses, robot home assistants, even robot wives and sexual partners - all this seems to be waiting for us in the near future. People are becoming less and less concerned with the idea of ​​the rise of evil and bloodthirsty machines and are increasingly seeking to make robots their friends. According to some estimates, by 2020 there will be about 100 million personal robots on the planet and “the next five years will lay the foundation for the fundamental changes that robots can bring to our homes and everyday lives.”

However, it is too early to make loud statements. These hundred million consist mainly of robot cleaners and automatic vacuum cleaners. But recent developments promise much more. Robots are already caring for residents in some nursing homes and becoming pets and household assistants in hundreds of families (this happens especially often in Japan and Western Europe). These are so-called social robots, or companion robots - machines that can imitate human or animal behavior.

Machines that can evoke sympathy and affection. Machines whose purpose is to brighten up human loneliness.

People in developed countries are rapidly aging and becoming increasingly lonely. For example, in Scandinavian countries, approximately 40% of households today consist of only one person. That is, in a typical Swedish apartment we are almost as likely to meet a single person as a family. People are increasingly choosing a single lifestyle on their own, preferring not to commit themselves to long-term relationships in order to find more space and time for themselves.

American sociologist Eric Kleinenberg noted in his book on this subject:

Masses of people decided to undertake this social experiment because, in their minds, such a life corresponds to the key values ​​of modernity - individual freedom, personal control and the desire for self-realization. Living alone gives us the opportunity to do what we want, when we want it and on the terms we set. Such an existence frees us from the need to take into account the demands and desires of our partner and allows us to concentrate on what is important to us.

But not everyone chooses this lifestyle voluntarily and independently. The development of hygiene and medicine has not only extended the lives of millions of people, but also made them more lonely. There are more and more elderly people, and increasingly there is no one to take care of them. Information technology makes it possible to communicate with anyone over a long distance, but increasingly forces us to remain in front of a screen instead of communicating with each other in the so-called “real world”. And social robots seem to be just another link in this chain. Will they be able to come to the aid of modern singles?

Social robots must be able to recognize and portray human emotions, adapt to human behavior, and change their own behavior patterns. Some robots can already be “educated”: they gradually change and form their character traits depending on how you treat them. They can be sociable and reserved, playful and calm. In addition, the robot must be able to distinguish one person from another. Relationships with the owner are one thing, and relationships with a random guest or passer-by are another.

Such a robot must be similar to a person, and people are terribly picky creatures. We consider the previous actions of those we know and treat them differently. If a toaster doesn't care about your personality when it toasts your breakfast bread, then a social robot is social to take into account the differences between you and everyone else. To do this, he will need a “theory of mind” - the ability to model a person’s mental state and predict his future actions.

In reality, robots are still very far from these skills. They can talk, plan your schedule, remind you to take medications or appointments, be your playmate or perform dances - sometimes much better than people can do. But you definitely won’t get sympathy from them. Robots don't feel anything.

Therefore, it would be more correct to say that we ourselves endow the robot with social qualities. Our psyche is designed in such a way that we easily take what seems to be real. We easily imagine a pink circle running away from a red triangle as a frightened innocent victim (see the famous Haider-Simmel experiment). It is, of course, even easier to endow intelligent machines with human qualities.

Scientists call this the natural tendency to anthropomorphize. Because of this tendency, for example, we imagine God as a gray-haired old man, believe in the evil eye, or attribute the most evil intentions to the alarm clock when it wakes us up the morning after a party. Recently, psychologists from the University of Chicago found that lonely people humanize animals and technological devices more often and more strongly than people who already have someone to talk to. The authors of the study note that such a human attitude towards “non-humans” can end badly:

Although anthropomorphization is one of the most creative approaches to satisfying the need for communication, it is difficult to form an intimate relationship with an inanimate object. This compensatory mechanism may prevent lonely people from taking risky, but potentially more rewarding, steps toward relationships with others.

Jennifer Bartz Professor of Psychology at McGill University

Psychologists who have studied the interaction of people with chatbots (for example, with the famous ELIZA, which imitates the behavior of a psychotherapist), have noted: when talking with robots, people strive to maintain the illusion of live communication, even if they know that they are essentially communicating with themselves. For example, ask the “right” questions so as not to drive the stupid bot into a stupor.

As for real rather than virtual robots, the relationship between artificial companions and elderly people in nursing homes has been best studied. In general, scientists agree that robotic companions improve overall well-being and reduce feelings of loneliness (see meta-analysis summarizing the results of 43 studies). In one comparative study, psychologists found that cute, furry robotic seals became the subject of conversation with other guests even more often than real dogs, and attracted more attention.

Maybe robots make us less lonely simply by making us interact more with each other?

But people who live not in nursing homes, but in their own apartments, also react positively to robotic companions. Moreover, they become very attached to them and sometimes prefer to communicate with the robot instead of interacting with people. Psychologists usually give a robot model to a volunteer who agrees to participate in a study for a few weeks, and at the end of this period they take it back. And here they often have to face difficulties.

The robot becomes “one of our own” for the person. He no longer wants to part with him.

Let's try to think: what is so seductive about communicating with a robot? They are, after all, lifeless and don’t feel anything! It's not just a matter of anthropomorphization. Robots, among other things, are more obedient and unobtrusive. They will not burden you with their problems when you have enough of your own, but will simply make coffee or give themselves a hug. And it’s not so important that these hugs will smell a little like silicone. You are always vulnerable in front of people, but in front of an insensitive robot you can be yourself.

Technology can be seductive if it frees us from feeling vulnerable. As it turns out, we are indeed quite vulnerable. We are lonely, but we suffer from a fear of intimacy. Virtual connections and social robots can offer the illusion of camaraderie without the commitment of friendship.

But robots still cannot satisfy the fundamental human need for live communication, the need to go beyond one’s own self. They can be good helpers, good caregivers and pets, but they are unlikely to become true friends.

In Spike Jonze's Her, a lonely writer falls in love with a cutting-edge operating system named Samantha. They begin a relationship, but in the end the hero finds out that Samantha is more interested in communicating with other operating systems: people only interfere with their continuous development. And then people are left alone. We have to solve human problems ourselves, because robots have their own, more important things to do.

This article uses a still from the film “Ex Machina” (2015).