Scientists programs an artificial intelligence able to tell whether a person is neurotic, extrovert, agreeable, and conscientious based on observing his or her daily eye movements.
The team, led by Tobias Loetscher from the University of South Australia, used the said AI to track eye movements of 42 people while they go about their daily and usual activities inside a university campus. The AI was encoded with different eye movement characteristics and was subsequently paired with a proven set of questionnaires on the Big Five personality traits used widely in psychology.
The scientists, from institutions in Australia and Germany, were able to detect four out of the Big Five personality traits based on the more practical use of eye movements as compared to previous similar studies on human behavior which are commonly limited to laboratory observations.
Based on the study which was published in the journal Frontiers in Human Neuroscience on April 13, the scientists concluded that advance and innovative design of human-computer systems is an important step toward a more realistic understanding of human traits and behavior.
Personality Traits And Eye Movements
The study noted that previous scientific studies have already established the correlation of personality traits and people's gaze.
Through these preceding studies, it was established that people with similar traits tend to move their eyes in a similar manner. For example, optimists are more likely to avoid looking at photos suggesting negative experiences as compared to the pessimists. People with more open mind fixate and dwell on locations when watching abstract animations while curious people linger more in a naturalistic scene.
These prior studies, however, were conducted amid two methodological limitations: first, early studies studied gaze and personality only descriptively or through personality scores without actual human observations; second, these early studies, although with human participants, were conducted in laboratories.
The first limitation obstructed researchers to actually observed eye movements beyond gaze alone, hence failing to determine whether eye movements can be used to detect personality traits.
In the second limitation, participants only looked on images appearing on computer screens. This set up is far from real life situations.
Artificial Intelligence And Eye Movements
The study said that the programmed machine used by the scientists is at between 7 and 15 percent better in reading personality traits.
This is because the participants in the study were asked to use eye-tracking smart glasses as they go about their daily lives around campus aside from answering the questionnaires on the Big Five personality traits.
In the future, the machine learning algorithm may be programmed into smartphones, self-driving cars, and robots that assist people. With the algorithm, these technologies could offer more personal and customized support to the users.
The researchers, however, warned that the AI needs to be strictly regulated so corporations will not abuse the program. Furthermore, with present controversies surrounding data acquisition by various tech companies, researchers warned that consumers might not be comfortable with this new algorithm they have designed.