Is artificial intelligence (AI) capable of suggesting appropriate behavior in emotional load situations? A team from the University of Geneva (UNIGE) and the University of Bern (UNIBE) put the test of six generative six AI, including the chatgpt, using emotional intelligence (EI) evaluations typically designed for humans. The result: these AIS exceeded average human performance and could even generate new tests in record time. These findings open new possibilities for AI in education, coaching and conflict management. The study is published in Communications Psychology.
Large language models (LLM) are artificial intelligence systems (AI) capable of processing, interpreting and generating human language. Chatgpt generative AI, for example, is based on this type of model. LLMS can answer questions and solve complex problems. But can they also suggest emotionally intelligent behavior?
These results are paving the way for AI to be used in contexts that are believed to be reserved for humans.
Emotionally loaded scenarios
To find out, a team from UNIBE, Institute of Psychology and the Swiss Aphysian Sciences Swiss Center of UNIGE (CISA) submitted to six LLM (Chatgpt-4, Chatgpt-O1, Gemini 1.5 Flash, Copilot 365, Claude 3.5 Haiku and Deepseek V3) to emotional intelligence tests. ” We chose five tests commonly used both in research and corporate environments. They involved emotional load scenarios designed to evaluate the ability to understand, regulate and manage emotions, ” says Katja Schlegel, principal teacher and researcher in the Personality Psychology Division, Differential Psychology and Evaluation at the Institute of Psychology of UNIBE, and main author of the study.
For example: one of Michael’s colleagues has stolen his idea and is being unfairly congratulated. What would be Michael’s most effective reaction?
a) Discuss with the colleague involved
b) Talk to your superior about the situation
c) Reset your colleague
d) Steal an idea
Here, option b) was considered the most appropriate.
At the same time, the same five tests were administered to human participants. “In the end, the LLM achieved significantly higher scores: 82% of correct versus versus 56% for humans. This suggests that these AI not only understand emotions, but also understand what it means to behave with emotional intelligence,” explains Marcello Mortillaro, a senior scientist at the Swiss Swiss Center for affairs (CISA), who was involved in research in the research.
New tests in record time
In a second stage, scientists asked ChatgPT-4 to create new emotional intelligence tests, with new scenarios. These automatically generated tests were taken by more than 400 participants. “They tried to be as reliable, clear and realistic as the original evidence, which had taken years to develop,” explains Katja Schlegel. Therefore, LLMs are not only able to find the best response between the various options available, but also to generate new scenarios adapted to the desired context. This reinforces the idea that LLMs, such as Chatgpt, have emotional knowledge and can reason about emotions, ” Marcello Mortillaro adds.
These results are paving the way for AI to be used in contexts that are believed to be reserved for humans, such as education, training or conflict management, provided that experts use and supervise it.