Researchers from Zhejiang University have introduced a novel AI training method that leverages human brain signals to enhance model categorization, as detailed in a paper published in Nature Communications. The study reveals that while increasing model parameters improves recognition of specific objects, it does not enhance, and may even hinder, the understanding of abstract concepts. By integrating neural activity data, the team aims to align AI models more closely with human cognitive structures, improving their ability to generalize and categorize concepts. The proposed method demonstrated significant improvements in AI performance, particularly in tasks requiring abstract concept recognition. Experiments showed a 20.5% improvement in distinguishing abstract concepts with minimal examples, outperforming larger models. This approach challenges the prevailing trend of expanding model sizes, suggesting that structured cognitive alignment may be more effective than sheer scale in developing AI with human-like thinking capabilities.