第一财经

Ali AI Investment Record: Heavy Exposure to Core Links of the AI Industry Chain, with a Profit Increase of Over 5 Times in 3 Years

原文:阿里AI投资全纪录:重仓AI产业链核心环节 3年浮盈超5倍

Summary of Key Points

Over the past three years, Alibaba's strategic investments have shifted entirely towards the AI sector. The approach has evolved from focusing on acquiring companies to control consumer ecosystems to investing in early-stage, small-scale startups at the forefront of technology, with the goal of achieving mutual benefit rather than complete control. This strategy covers the entire AI value chain, including chips, large models, AI applications, and embodied intelligence. To date, Alibaba has invested approximately 36 billion yuan, resulting in a floating profit of over 210 billion yuan (an average return of 5.8 times the investment amount). Notable examples include investments in ChangXin Technology (17-fold return), Zhipu AI (70-fold return), and YueZhiDanMian (over 6-fold return). Alibaba not only makes money from these investments but also provides computing power through its Alibaba Cloud platform, becoming a vital source of support for China's AI industry.

I. Alibaba’s AI Investment Performance: 210 Billion Yuan in Floating Profit from a 36 Billion Yuan Capital Outlay

Alibaba’s AI investment efforts have been incredibly profitable:

  • ChangXin Technology: With a total investment of 7.6 billion yuan, the company’s equity value has increased to about 130 billion yuan post-listing, representing a 17-fold return (equivalent to a 100,000 yuan investment turning into 1.7 million yuan).
  • Zhipu AI: A single investment has generated a return of 14 billion yuan, a 70-fold increase.
  • MiniMax: An investment of 2.9 billion yuan resulted in a return of 18 billion yuan, a more than 6-fold increase.
  • YueZhiDanMian: Although not yet listed, the floating profit exceeds 40 billion yuan.
  • Overall: With a starting capital of 36 billion yuan, Alibaba has now earned a total of 210 billion yuan in profits. Moreover, there are many other companies that have not yet gone public (such as YuanLiLingJi in embodied intelligence), and their future returns are expected to rise even further.

II. Comprehensive AI Chain Coverage: From Chips to Robot Applications

Alibaba’s AI investments are well-structured along the industry chain from hardware to software to applications:

  • Chips (computing power foundation): Investments have been made in ChangXin (domestic memory chips for replacing imported ones), LanQi (server interconnection chips), as well as Aojie and XiZhi. Some of these companies are already listed, while others are in the process of going public.
  • Large models (AI brains): Alibaba invested in Zhipu AI, MiniMax, and YueZhiDanMian at the beginning of the development of domestic large models in 2023, covering all the leading players. The first two companies have already gone public, and YueZhiDanMian also has a high valuation.
  • AI applications (practical use cases): Investments include Genspark (a general AI assistant), Dify (an open-source tool), and HaiYi (which is expanding into the international AIGC market), covering various everyday AI tools.
  • Embodied intelligence (robots): Alibaba has invested in ZhujiDongLi (humanoid robots), YuShu Technology (quadruped robots set to go public on the A-share market), YuanLiLingJi (robot “brains”), and even in brain-computer interface companies like JieTiYiZhi, extending AI from the virtual to the physical realm.

III. Major Change in Investment Strategy: From Controlling Companies to Collaborating for Mutual Benefit

Alibaba’s investment approach has changed significantly. It no longer aims to acquire control of companies but to work together with startups for mutual growth:

  • No desire for control: Alibaba holds only a small stake in its investees (for example, 5% in ChangXin before its listing, allowing the founders to make decisions independently).
  • Early-stage and small-scale investments: Alibaba enters industries when they are still nascent (e.g., investing in ChangXin during a downturn in 2021 or when the prospects for large models were uncertain in 2023), providing funding and resources to early teams.
  • Mutual benefit model: If an investee needs computing power, Alibaba Cloud offers it (for example, YueZhiDanMian spent 200 million US dollars on Alibaba Cloud services); if an investee wants to sell products, Alibaba Cloud’s platform helps with promotion (e.g., Zhipu AI sold its models through Alibaba Cloud). This approach allows both parties to profit.

IV. Why Alibaba Can Become a Leader in the AI Industry

Alibaba’s success in AI investments is not due to luck but to its strong capabilities:

  • Financial resources: By selling off businesses like Yintai and GaoXin Retail and reducing its stake in companies like Bilibili and YTO Express, Alibaba has freed up funds to invest in AI, with plans to allocate another 380 billion yuan for AI infrastructure development.
  • Computing power: Alibaba Cloud is the largest AI cloud provider in China, providing much-needed computing resources to its investees (for example, MiniMax uses Alibaba Cloud as its primary cloud service provider).
  • Insight and vision: Alibaba dares to take risks when others are hesitant (e.g., investing in ChangXin in 2021 when domestic memory technology was still under question) and focuses on leading companies (top three in large models and early-stage startups in embodied intelligence).

V. The Implications of This Change for Alibaba and China’s AI Industry

  • For Alibaba: It marks a transition from a consumer giant to a technology powerhouse, with improved capital efficiency (investing profits from traditional businesses into AI, resulting in multiple-fold returns). Alibaba Cloud’s growth rate has reached 45%, providing a solid foundation for future expansion.
  • For China’s AI industry: Alibaba acts as a support system, providing funding and computing power to early-stage companies, enabling them to thrive and grow (for example, ChangXin becoming a major player in domestic memory production, and Zhipu AI becoming the first Chinese company listed in the large model sector). Large companies no longer compete for market share but rather support the industry’s development, which is crucial for overcoming technological barriers in China.

In summary, Alibaba’s AI investments have not only been profitable but have also significantly contributed to the growth of China’s AI industry, representing a win-win situation. The listing of ChangXin is just the beginning, and many more investees are set to go public, with Alibaba’s AI story still unfolding.