Stanford University researchers have created AI agents that accurately forecast human behavior. Dr. Joon Sung Park and his team conducted a study showing that a two-hour interview provides sufficient data for AI to mimic human decision-making patterns with 85% accuracy. These digital clones, known as LoRAs, go beyond deepfakes. They can be used to profile users and test their responses to various stimuli. The research involved 1,052 Americans from diverse demographics engaging in interviews exploring life stories, values, and social perspectives. Unlike other studies, researchers used an 'expert reflection' module to analyze interview data from multiple professional perspectives. The AI agents, powered by GPT-4o, successfully replicated human behavior in controlled scenarios. The system demonstrated reduced bias across racial and ideological groups compared to traditional approaches. While the technology shows promise, ethical considerations regarding the use of AI clones in employment settings are crucial as the line between human and digital identity blurs. Read more AI-generated news on: https://app.chaingpt.org/news