With DeepSeek launching its large language models V3 and R1, the global AI landscape is becoming transformed. Developed in China, these models perform well with high cost-effectiveness. Through its open-source approach, DeepSeek has reduced the entry barriers for AI application. The success of DeepSeek demonstrates not only China's leading role in the global democratization of AI, but also a vision of openness and collaboration.

Recently, Clément Delangue, co-founder of Hugging Face, the world's largest open-source platform, called DeepSeek-R1 the most popular model on the Hugging Face platform. DeepSeek-R1, with numerous modified derivative models created from it, has been downloaded more than 10 million times. This illustrates DeepSeek's technological capability and popularity, and the demand within the open-source community for efficient, cost-effective AI solutions.

DeepSeek's open-source nature exemplifies China's open cooperation globally. Unlike some closed ecosystems, DeepSeek embraces technological democratization.

The training cost of DeepSeek-V3 is approximately USD 5.57 million, much lower than ChatGPT-4's billion-dollar cost. By being open source, DeepSeek has accelerated innovation and brought significant benefits to emerging markets.

Indonesia's educational company, Ruangguru, is an example. Its AI-driven robot products help students to address study challenges. Through DeepSeek, Ruangguru has made its chatbot more effective. It can answer various questions, retaining more user learning history at a much lower cost.

"In the past, our AI might have only been able to remember two to three paragraphs of learning context due to high costs. However, with DeepSeek, AI can remember more learning history because it is cheaper. This is very useful because when students interact, AI can provide recommendations for appropriate learning materials and methods," Alvin Francis Tamie, the company's chief technology officer, said during a panel discussion with East Ventures, a leading venture capital firm.

This illustrates how Chinese software, by being open source to the world, has broken traditional barriers, facilitating technological progress and contributing novel solutions to challenges.

Additionally, the success of DeepSeek has boosted global technology sovereignty. Distinct from the closed systems and monopolies of Western AI technologies, it is promoting technological democratization.

Besides Chinese enterprises and global tech giants, enterprises like Russia's Yandex and India's Yotta Data Services have also developed localized models under the DeepSeek framework. This signals a shift from a "giant monopoly" to a "multipolar symbiosis" for AI, indicating that open source not only enables technological breakthroughs but also drives balanced global technological development.

By making its research and its weighting publicly available, DeepSeek has set an example for ethical governance in AI.

As Ambuj Tewari, professor of statistics at the University of Michigan and a distinguished expert in AI and machine learning, stated in an interview with University of Michigan News, the DeepSeek model can be run on servers owned by U.S. companies as its weights are publicly accessible.

In addition to reinforcing public trust in AI technology, such efforts also contribute novel ideas to global AI regulation. The open-source model has significant transparency and accountability, an essential requirement to build a fairer digital world.

Finally, DeepSeek's open-source strategy is also impacting traditional technological hegemony while shaping a new industrial ecosystem. In China, its models are being integrated in various industries like telecommunications, automotive, securities, and education.

By 2025, its intelligent computing power is estimated to exceed 1,037 EFLOPS in China, with open-source models as a major driver. This demonstrates that AI promotes the intelligent upgrade of all sectors, with open source as a transformative force for social progress and global cooperation.

Major international players like Microsoft, AWS and NVIDIA have integrated DeepSeek into their ecosystems rapidly, marking a shift from theoretical discussion to implementation. This highlights the substantial impact of the open-source model globally.

At the recent AI Action Summit in Paris, over 60 signatories released a statement on inclusive and sustainable AI for people and the planet. China was one of the signatories, and not the U.S. and the UK. This shows the scant international support for the self-centered and exclusive logic of AI hegemonic development.

In contrast, China is widely recognized for its open, inclusive, equitable, and beneficial AI governance philosophy.

The success of DeepSeek illustrates that sharing is the true value of innovation. China is playing a vital role in the democratization of global AI by making core technologies open, advancing AI applications worldwide, and providing valuable lessons in addressing various challenges.

This marks a turning point in the history of digital civilization, in which the shared future of humanity is collectively created by three forces: technological accessibility, collaborative innovation and ethical governance.