DeeCamp2021, a global artificial intelligence training camp co-hosted by Sinovation Ventures and Tsinghua University's Institute for AI Industry Research (AIR), concluded on August 17, 2021, with a series of notable achievements. The camp, now in its fifth year, is a philanthropic initiative aimed at cultivating applied AI talent among college students. This year, the program attracted 200 top students from prestigious universities around the world, forming 39 teams to tackle 13 challenges across four tracks: language, healthcare, innovation, and business. After a rigorous process of project conception, development, and testing, 35 teams presented innovative AI demos, with six teams advancing to the finals. Among these, the GeneBERT project emerged as the "double champion," winning both the annual grand prize and the healthcare track prize, alongside three other teams that won their respective track championships. The GeneBERT project, led by a team of international scholars with expertise in both life sciences and computer science, focused on developing a deep pre-training model for gene diagnosis. The team, comprising Mo Shentong from Carnegie Mellon University, Fu Xi from Columbia University, Hong Chenyang and Chen Yizhen from the Chinese University of Hong Kong, and Zheng Yuxuan from East China Normal University, worked under the guidance of Professor Yan Yan from Tsinghua University's AIR. Their project addressed a critical gap in the current market, where most gene testing products only cover a small fraction (0.04%-0.6%) of the entire genome, primarily focusing on single-gene congenital disorders. By expanding the modeling scope to include functional non-coding regions and incorporating various interactions within non-coding regulatory areas, the GeneBERT team provided a new paradigm for analyzing gene mutations in non-coding regions. This innovation holds significant potential for drug development, basic medical pathology research, and personalized treatment plans, with applications ranging from newborn disease screening to cancer diagnosis and therapy. Wang Yonggang, the Executive Dean of Sinovation Ventures' AI Engineering Institute, praised the GeneBERT team for their innovative spirit and ability to apply cutting-edge AI algorithms to life sciences. Mo Shentong, the team captain, expressed his desire to collaborate further with academic partners and industry experts to develop a comprehensive service from gene testing to personalized treatment design. Fu Xi added that DeeCamp provided valuable insights into the commercialization of research and the societal demand for new medical technologies, inspiring the team to strive for more complete academic and commercial outcomes in the future. In addition to GeneBERT, three other teams won their respective track championships. The Free Lunch team from Nanjing University won the "AI-driven business scenarios" track with their project "Golden Finger," which uses AI to enhance financial data analysis by extracting events and related entity information from large volumes of financial texts. The Edutainment Club team, comprising students from Tsinghua University, University of Macau, Northwest A&F University, and Columbia University, won the "AI innovation and creativity" track with their educational board game "Click," which combines a recommendation system with a STEM question bank to make learning more engaging and effective. Lastly, the Little Yellow Duck Team 2, from Tongji University and Boston University, won the "intelligent innovation in language and knowledge" track with "AI Recommend," a system that generates marketing copy tailored to different audience styles based on product features or images. Dr. Kai-Fu Lee, the Chairman and CEO of Sinovation Ventures, emphasized the importance of AI engineers understanding the ethical and moral implications of their work to avoid negative impacts on users. He highlighted the need for fairness and explainability in AI applications, particularly in sectors like healthcare, where errors can have serious consequences. Professor Ya-Qin Zhang, the Director of Tsinghua University's AIR, noted the potential of AI and automation in the B2B sector, suggesting that students could focus on verticals such as enterprise services and industrialization to solve real-world problems. He commended the GeneBERT team for their interdisciplinary approach and commercialization strategy, which set a good example for future participants. Over the past five years, DeeCamp has grown from a small experimental camp to a large-scale training program, training over 1,300 students annually. The program has received support from numerous top scholars and experts, including Dr. Kai-Fu Lee, Professor Ya-Qin Zhang, Andrew Ng, Hongjiang Zhang, Zhizhong Zhou, Ming Zhou, Weiyong Ma, and Tong Zhang, who have provided lectures on AI developments, future opportunities, and challenges. These sessions cover a wide range of topics, from machine learning and computer vision to the commercialization of AI products and investment opportunities in the AI era. DeeCamp has also garnered support from many companies, which have contributed valuable resources, including topics, case studies, and mentors. For instance, Huian Jinkē has used DeeCamp to help students apply NLP and knowledge graph technology to uncover corporate fraud risks, while Refinitiv has collaborated with DeeCamp for three years to explore further commercial and technical cooperation in fintech. Guokr, a science communication platform, has supported DeeCamp by connecting academic resources with industry needs, aiming to help companies discover new technologies and assist future technical elites in understanding the industry better. The success of DeeCamp is evident in the career paths of its alumni. Many have gone on to work in leading tech companies such as Huawei, Tencent, ByteDance, Kuaishou, and Megvii, while others continue to pursue research in academic institutions. Some alumni have even started their own companies, leveraging the skills and ideas developed during their time at DeeCamp. The overarching goal of DeeCamp is to equip students with the technical, practical, and commercial skills needed to address real-world challenges and transform AI algorithms into successful products and solutions. As the AI industry continues to evolve, DeeCamp aims to play a pivotal role in shaping the next generation of AI professionals who will contribute to the advancement of technology and the betterment of society.