北大团队最新研究：AI零基础自主推导出牛顿第二定律

Core Insights - The AI system developed by Peking University, named "AI-Newton," has demonstrated the ability to independently derive classical mechanics laws from experimental data, marking a significant shift in the role of AI in scientific research [1][2] - This advancement indicates a transition of AI from being a mere tool to becoming an independent cognitive entity in scientific inquiry [1] Group 1: AI System Capabilities - "AI-Newton" was tested with 46 sets of experimental data without any pre-set physical knowledge, successfully identifying motion patterns and constructing fundamental concepts like "force," "mass," and "acceleration" [1] - The system ultimately derived Newton's second law, F=ma, using a reasoning strategy akin to the scientific method of hypothesis and verification, rather than simple fitting as seen in traditional large models [1] Group 2: Research Implications - The programming approach of "AI-Newton" is expected to encourage models to actively approach core concepts, potentially becoming a crucial method for future scientific discoveries [2] - The research team plans to extend this methodology to more complex areas, including exploring patterns within quantum systems, indicating a shift of AI from data analysis to being a participant in theory generation [2]