Israeli Researcher’s AI Identifies ‘Social Norm Violations’
An Israeli researcher at Ben-Gurion University of the Negev has developed an artificial intelligence (AI) system capable of identifying social norm violations in written text. Professor Yair Neuman, with support from engineer Yochai Cohen and funding from the US military, built the system using GPT-3, zero-shot text classification, and automatic rule discovery.
The system was trained to recognize 10 social emotions, including competence, politeness, trust, discipline, caring, agreeableness, success, conformity, decency, and loyalty. It successfully classified texts into these 10 groups and determined whether they were positive or negative.
To validate the models, the system was tested on two large datasets of short texts, and the results proved its accuracy. The findings were recently published in the journal Scientific Reports, highlighting the potential of this technology.
The development of this AI system comes as part of the US Defense Department’s Defense Advanced Research Projects Agency (DARPA) initiative, called The Computational Cultural Understand program. The program aims to create cross-cultural language understanding technologies to enhance situational awareness and interactional effectiveness. Cross-cultural miscommunication can disrupt negotiations and even contribute to conflicts.
Professor Neuman, the head of The Functor Lab in the Department of Cognitive and Brain Sciences at BGU, stated that while this is preliminary work, it provides strong evidence that their approach is correct and can be expanded to include more social norms.
This breakthrough in AI technology has the potential to streamline communication processes and prevent misunderstandings in various domains. By identifying social norm violations in text samples, the system could be utilized in diplomacy, negotiations, and conflict prevention efforts.
It is worth noting that as AI continues to advance, ethical considerations must be taken into account. The system’s ability to classify texts based on social norms raises questions about its potential impact on freedom of expression and cultural relativism. Striking a balance between the benefits of such AI technology and the preservation of human rights and diversity will be crucial moving forward.
Overall, Professor Neuman’s AI system represents a significant step forward in understanding and analyzing social interactions through the lens of text analysis. As the technology continues to evolve, it has the potential to revolutionize communication and contribute to more effective cross-cultural understanding.
