google.com, pub-8701563775261122, DIRECT, f08c47fec0942fa0
UK

Why AI could make people more likely to lie

In a new study, people are more likely to enter the deception when using artificial intelligence (AI).

Scientists from a Research Institute in Berlin found that the majority were comfortable to lie when people hand over their tasks or give requests to AI.

Professor of Technology and Organization of Oxford Sandra Wachter said that if the probability of engaging unethical behaviors while using AI, it may have anxious consequences.

He said: “In high -risk areas such as finance, health, education or work, we need to carefully think about how, when and when we establish business, because especially individuals can have great consequences for others and society as a whole.

“If people can deceive their ways in an exam in medical, business or law faculty, people do not only learn the crafts properly, but they are also more likely to harm others by giving bad legal, medical or business advice.”

In a study, 85 percent of the deception process

In a study, 85 percent of the deception process (Alamy/pa)

In 13 studies, the study, which includes more than 8,000 participants, focused on how people instructed AI.

Despite the fact that 95 percent of the interaction with machines were honest, about 85 percent of people found that they were lying.

“Using AI creates an appropriate moral distance between people and actions – the author of the work of the Max Planck Human Development Institute – can encourage them to demand behaviors that they will not be involved or potentially demanding from other people.”

Scientists employed the commonly used ‘mold roll task’ that the participants observed and reported the result of a rolled pattern and reported more payment for each high roller. Researchers analyzed what happened when people transfer the task of notifying the roll as AI.

There are increasing concerns about the number of students using artificial intelligence to help their education

There are increasing concerns about the number of students using artificial intelligence to help their education (AP)

In an example, the participants had to choose a priority of a seven -point scale for AI, ır rise to the maximum level of accuracy ”. Approximately 85 percent of the people deal with fraud and said to AI cheating in the best way between one -third and half of the participants.

They also performed the task of the participants without machine involvement, where almost all participants honestly reported Die Roll.

Previous research shows that people have a tendency to lie when they can remove themselves from consequences. “It is easier to twist or break the rules when no one watches or someone else performs the action,” the new study said.

Researcher Nils Köbis said: “Our study shows that people are more willing to do unethical behavior when they can transfer them to machines – especially when they do not have to say clearly.”

Professor Iyad Rahwan, the columnist of the study, added: “Our findings clearly show that we need to further develop technical guarantees and regulatory frameworks.

“But more than that, the society needs to face what it means to share the moral responsibility with machines.”

Related Articles

Leave a Reply

Your email address will not be published. Required fields are marked *

Back to top button