To The Tune Of Artificial Intelligence: Why Children Trust Robots More Than Themselves - Alternative View

Table of contents:

To The Tune Of Artificial Intelligence: Why Children Trust Robots More Than Themselves - Alternative View
To The Tune Of Artificial Intelligence: Why Children Trust Robots More Than Themselves - Alternative View

Video: To The Tune Of Artificial Intelligence: Why Children Trust Robots More Than Themselves - Alternative View

Video: To The Tune Of Artificial Intelligence: Why Children Trust Robots More Than Themselves - Alternative View
Video: Необычный случай с Алексом Льюисом (документальный фильм о чудесном чуде) - Реальные истории 2024, November
Anonim

Scientists from the University of Plymouth have shown that children believe everything that artificial intelligence says.

Artificial intelligence can now be found in almost every iron, and even more so in children's toys. Today's kids learn to use gadgets before they read, especially since they don't even need to know a letter if they have a voice assistant. But any technical failure or vulnerability in the system can make such a device quite dangerous if children perceive everything that the robot says as absolute truth.

Therefore, scientists from the University of Plymouth (UK) decided to look at how children interact with their smart toys and whether they can critically perceive the "words" of robots.

Robots know better

To see if children trust artificial intelligence, scientists repeated an experiment by renowned psychologist Solomon Asch. In 1951, Asch conducted a series of experiments on social conformism. Under the guise of an eye test, he suggested that people look at four lines and choose two of the same length. Some of the people in the group were dummy actors who gave obviously wrong answers. In total, the psychologist conducted a series of 20 such experiments. As a result, it turned out that almost 75% of the subjects at least once adjusted to public opinion, albeit incorrect.

Now the researchers have modified the experiment. Instead of dummy actors, the wrong answers were prompted by robotic toys. And it turned out that robots can confuse children 25% of the time. Despite the fact that on their own children found the correct answers much more often - in 87% of cases.

“While the children were alone in the room, they were doing quite well at the tasks,” said one of the study's authors, Tony Belfime, a robotics specialist at the University of Plymouth. “But as soon as the robots appeared and began to give the wrong answers, the children immediately followed them.

Promotional video:

AI adults don't believe

In the second part of the experiment, the researchers decided to observe how exactly children react to prompts from robots. This time, the kids were offered tasks in which it seemed impossible to make a mistake - the correct answers are obvious! Even so, 74% of the subjects first watched what the robot would say, and then repeated its answer word for word. A total of 43 children participated in the test.

“People tend to trust machines, this is a kind of bias,” says Alan Wagner, an aerospace engineering specialist at the University of Pennsylvania (USA). - We are used to trusting robots and gadgets because we believe that they are more knowledgeable than we are.

However, adults in a similar experiment turned out to be far from being as trusting as children. The scientists repeated the test with 60 adults, and their answers were not influenced by the robot's prompts.

“Perhaps the fact is that these robots looked like children's toys, and their answers did not inspire confidence in the adult audience,” says Tony Belfime. - Probably, if the robots looked more solid, or if Siri (Apple's voice assistant - ed.) Gave prompts, the result would be different.

The adult participants in the experiment themselves later admitted that they decided that the robots had simply broken down. Well, or that these tasks turned out to be too difficult for artificial intelligence.

Eventually

It is clear that children at this age were let down by the skill of critical thinking - it is formed only with age and experience. However, in the case of AI and smart toys, such gullibility turns out to be dangerous, the researchers warn. Such technologies are increasingly being used in education and for working with children. Moreover, robots are increasingly being given human features - from hands, feet, heads to voice. By the way, this is why female voices are often used in voice assistants - this increases user confidence.

EXPERT COMMENTARY

Elena Smirnova, Doctor of Psychology, Head of the Center for Psychological and Pedagogical Expertise of Games and Toys, Moscow City Psychological and Pedagogical University:

- The vast majority of smart toys are now singing, moving, talking by themselves. All this activity paralyzes the child's own activity. It's fun for adults to look at such a toy. But it is impossible for a child to play with her. If the toy speaks by itself, it cannot be assigned its own voice in the imagination. If she does something herself, it is difficult to integrate her into some game plots. Smart toys and robots are self-sufficient and closed, the child cannot put his feelings and emotions into them. Therefore, it is not worth saying that such toys develop. Rather, they are just marketing gimmicks.

KSENIYA KONYUKHOVA