Can artificial intelligence detect lies?

[zakiyatasnim]

Personal computers are not very good at detecting distortions, and this has become a problem as technology is increasingly used to solve important problems. Scientists from the University of Southern California have learned whether artificial intelligence can detect lies.

Algorithms typically fail when used as lie detectors, a research group studying emotion recognition has found. Two recently completed experiments completely disregard the foundations of popular psychology and artificial intelligence training methods that rely on facial expressions reflecting thoughts.

“Both people and so-called emotion-reading algorithms quote benin number data the benin number data folk wisdom that our feelings are written on our faces. This is far from the truth. People smile when they are angry or upset, masking real anxiety, and often facial expressions have nothing to do with the inner state of the soul, but only show the cultural traditions of a particular people,” notes Professor Jonathan Gretch.

Of course, people know that people around them can lie with a casual expression on their faces. Poker players bluff. Employers are deceived by job seekers, and an unfaithful husband lies to his wife. Politicians can also cheerfully promise things they can never fulfill.

However, algorithms are not very good at exposing human deceit, although they are increasingly used to read human emotions and make fateful decisions. For example, the Department of Homeland Security is investing in programs to predict potential threats to society. Some countries use mass surveillance to monitor the communications of the population. Algorithms are used in focus groups, marketing campaigns, and when screening potential employees.