AI for Science 爆发：AI 能解锁伟大的科学发现吗？ | S10E29

AI for Science is fundamentally transforming life sciences by enabling the creation of "virtual cells" and biological world models. Song Le, CTO of GenBio AI, explains that while AI excels at logical deduction, data processing, and automating research workflows, the ultimate challenge remains the creative leap required for scientific breakthroughs. By integrating multi-scale data—from DNA and proteins to entire cellular systems—AI models can simulate drug responses and disease mechanisms, significantly reducing the need for costly, time-consuming physical experiments. As these models evolve, they are shifting from simple predictive tools to sophisticated simulators that allow researchers to plan experiments more efficiently, marking a transition toward an AI-native era in drug discovery and biological research. This evolution mirrors the scaling laws seen in large language models, where increasing data quality and model complexity drives deeper, more accurate scientific insights.