Core Insights - The article discusses the emergence of AI for Science (AI4S) as a transformative force in scientific discovery, particularly in addressing critical challenges in foundational science [2][6][36] - The company DeepMind and the Baker team developed a deep learning model called "RFdiffusion," which predicted approximately 200 million protein structures, showcasing the potential of AI in scientific research [2][36] - The founder of Deep Origin, Jia Haojun, emphasizes the importance of AI in enabling new scientific discoveries rather than merely serving as a tool for existing processes [2][36] Company Overview - Deep Origin was founded by Jia Haojun, who integrated generative AI with first principles to apply AI in material research [3][33] - The company has developed six proprietary algorithm modules, which are integrated into a self-developed platform named "ReactiveAI" [3][33][46] - The platform has recently been upgraded to a material discovery agent called "Agent Mira," which autonomously mobilizes data and resources for chemical material development [4][34] Industry Trends - In 2025, AI4S is expected to reach a critical turning point, with significant investments and initiatives from both the U.S. and China aimed at leveraging AI for scientific research [6][36] - Major companies like Tencent, Alibaba, and ByteDance are rapidly establishing AI4S teams and initiatives, indicating a strong competitive landscape [6][36] - The "Artificial Intelligence+" plan in China highlights AI4S as a key direction for upgrading scientific discovery paradigms [6][36] Funding and Growth - Deep Origin completed a Series A financing round exceeding 100 million RMB, led by Alibaba's entrepreneur fund and Ant Group, among others [7][37] - The company has received multiple rounds of funding, totaling several hundred million RMB, reflecting the growing interest in the AI4S sector [52][56] - The AI4S sector has become a favored area for investment, with notable companies achieving significant funding and market presence [52][56] Commercialization Efforts - Deep Origin is actively expanding its client base, securing contracts worth millions in various industries, including beauty and materials energy [24][56] - The company successfully collaborated with a European beauty giant to address stability issues in active ingredients, demonstrating the practical value of its AI platform [26][56] - The approach of "co-developing with clients" is seen as a more effective way to popularize AI applications compared to merely selling the platform [26][56] Technical Innovations - Deep Origin's unique "ECML system" combines AI model predictions, computational support, and experimental validation, significantly enhancing computational efficiency [45][46] - The company has developed a layered generation architecture to ensure the physical feasibility of generated material structures [45][46] - The integration of specialized algorithms tailored for chemical reactions and material performance prediction creates a competitive edge that is difficult to replicate [29][46]