CITIC SEC: Decoding OpenAI Founder Sam Altman's AI Investment Landscape.

CITIC SEC released a research report stating that the investment focus of OpenAI founder Sam Altman in AI includes: horizontally expanding multimodal and cost-effective small models, vertically laying out upstream fundamentals such as AI chips, nuclear energy, and humanoid Siasun Robot & Automation, cameras, and other AI terminals. In addition, the top five application directions for Altman's layout are enterprise services, efficiency tools, financial technology, education and training, and healthcare. The investment opportunities in China's AI development mainly include: horizontally focusing on enterprises with ample cash reserves and early implementation of cost-effective models; upstream focus on domestic alternatives for AI chips, databases, as well as new infrastructure developments such as nuclear power and AIDC; downstream focus on intelligent terminals such as humanoid robots and cameras, and applications landing in enterprise services, finance, education, and other fields. Altman's investment landscape overview: Altman has invested in over 400 projects, with AI being a key investment area. The rapid development of generative AI has attracted high global investment growth, with the United States being in an absolute leading position. According to Stanford, private investment in the AI field in the United States in 2023 is 8.7 times that of China. OpenAI is a leading US company in the global generative AI development, and its founder Sam Altman is also an outstanding investor. According to Wall Street, as of early 2024, Altman and his venture capital fund have invested in over 400 companies, with a share value exceeding $2.8 billion, with AI being their key investment area. According to CB Insights, SVTR AI, and various company websites, Altman and his investment platform have invested in approximately 92 projects since 2019, with 82 of them being related to AI. Among the 82 AI-related investment projects, 17 are related to the AI infrastructure layer, 3 are related to AI terminals, and 62 are related to AI applications. Altman's investment directions summarized: horizontally expanding multimodal and small models, and vertically laying out the AI industry upstream and downstream. 1) Horizontally, multimodal products: Since launching GPT in 2018, OpenAI has continued to expand its layout with products such as DALLE, Sora, and GPT-4o, which are likely to accelerate the practical implementation of AI in complex real-world scenarios. High-cost-effective small models: In July 2024, OpenAI launched the small model GPT-4o mini with a parameter scale of about 7 billion, realizing higher cost-effectiveness by reducing model size and performance. Brand new inference model: On September 12, 2024, the o1 model was launched, to improve performance by introducing internal thinking chains and complex inference processes. 2) Upstream: Layout of AI infrastructure. In the context of chip shortages, substantial consumption of resources such as electricity, and the increasing importance of data, Altman has invested in important targets such as AI chip company Rain Neuromorphics, nuclear fusion companies Oklo and Helion Energy, real-time analytics database Rockset, to explore AI infrastructure solutions. 3) Downstream: Layout of AI terminals and applications. In terms of terminals, Altman has invested in companies such as Humane of AI Pin, Figure AI of humanoid Siasun Robot & Automation, and Opal of high-end network cameras, mainly targeting the fields of human-machine interaction, labor replacement, and household consumption. In terms of applications, enterprise services, efficiency tools, financial technology, education and training, and healthcare are the top five application areas Altman has laid out in, with investments in companies such as Induced AI (workflow automation), qqbot (programming development tools), and Slope (B2B payment platform). The mapping of Altman's investments in China: Fierce competition in domestic model development, steady progress in upstream infrastructure, and downstream applications. 1) Horizontally: Intense competition among domestic models, unicorn companies emerging. China currently has 117 registered domestic large-scale models (Cyberspace Administration, as of March 2024), with companies such as Zhipei, Dark Side of the Moon, Zero One Electronics, Bai Chuan Intelligent, and Minimax being the "Five Tiger Models", with Douyin's "Doubao" model's performance leading by a large margin. 2) Upstream infrastructure: Active layout in algorithm construction, with accelerated development of domestic alternatives. AI-driven domestic cloud provider capital expenditures have greatly increased, with a 124% year-on-year increase in combined capital expenditures by BAT in the first half of 2024, reflecting the growing demand for infrastructure. The replacement of AI chips and databases by domestic products is accelerating; the development of AIDC, nuclear power, and other infrastructure is picking up pace. 3) Downstream terminals and applications: In terms of terminals, the capital market enthusiasm for humanoid Siasun Robot & Automation has surged, with funding reaching 5.5 billion yuan in 23 years (IT Orange), products such as Yushu G-1 and Zhiyuan Expedition A2 gradually coming online; Omedia predicts that the penetration rate of intelligent cameras in China is expected to increase from 15% to 63% from 2019 to 2024. In terms of applications, BEISEN HOLDING launched the "Beisen AI Family", Alibaba launched the "Tongyi" series, and Hithink RoyalFlush Information Network launched the "Ask Finance HithinkGPT Large Model", indicating active progress in domestic AI applications. Risk factors: Risks of technological innovation falling short of expectations; Risks of commercialization progress of AI applications falling short of expectations; Risks of intensified competition in popular sub-directions of the industry; Risks of policy support falling short of expectations; Risks of macroeconomic fluctuations leading to lower-than-expected downstream demand; Risks of increasing international conflict limiting domestic AI development; Risks of potential ethical, moral, and user privacy issues in AI.