A new Athics webinar on Monday 24 March at 5pm
Personality analysis with AI is one of the most fascinating applications of artificial intelligence, starting from the ability of generative models to estimate personality traits through the analysis of written language. A study recently published in Frontiers in Artificial Intelligence entitled “On the emergent capabilities of ChatGPT 4 to estimate personality traits” investigated precisely this aspect, highlighting the potential and limitations of ChatGPT 4 in recognizing people’s psychological traits based on the Big Five model.
Personality analysis with AI is a field very close to Athics. In fact, Portrait technology is based precisely on the ability to detect the main psychometric traits that emerge from the frequency of use of function words within a written text.
For this reason, we enthusiastically welcomed the academic attention for this particular area of application of artificial intelligence and invited Prof. Marco Piastra of the University of Pavia and Prof. Patrizia Catellani of the Catholic University of Milan to present the recent publication in an Athics webinar, scheduled for Monday 24 March at 5 pm, live on Linkedin.
Personality Analysis with AI: the potential of ChatGPT 4
The study on personality analysis with AI, conducted by the Department of Industrial, Information and Bioengineering Engineering of the University of Pavia and the Department of Psychology of the Catholic University of the Sacred Heart in Milan, analyzed the ability of ChatGPT 4 to predict personality traits through a zero-shot approach, that is, without specific training on this task. Using two datasets of written texts, the researchers compared the AI estimates with the authors’ self-assessments. The results show a moderate correlation between the model’s predictions and the actual psychological characteristics of individuals, suggesting that Large Language Models (LLM) could have a role in automated psychological analysis, albeit with some limitations.
Athics Webinar “Understanding people with AI”
During the webinar, the starting scenario and the motivations behind the study, the expected results, the possible applications of personality analysis with AI, the future implications and the possible further developments of this study will be examined.