International Struggle: China and the United States Compete to Mold the Future of Artificial Intelligence
The world of artificial intelligence (AI) is evolving rapidly, with the latest developments making significant strides in advancing the technology. One of the key focus points in global AI development is the latest AI model, which is expected to contribute significantly to the field.
DeepSeek Project
In January 2025, Chinese AI startup DeepSeek released its R1 model. This model has made waves in the AI community, as it competes with leading US models, such as those from OpenAI and Anthropic, but at a much lower cost and with significantly less computing power required. This challenge to the assumed dominance of US AI companies has raised questions about the effectiveness of US export controls on cutting-edge chips.
DeepSeek's model innovations focus on algorithmic efficiency rather than brute computational scale. Techniques such as mixed-precision arithmetic and a "mixture of experts" approach are used, which activate fewer parameters at inference time, significantly reducing energy and operational costs. The open-source nature of DeepSeek's models allows researchers worldwide to freely access, fine-tune, and build upon them, fostering a global ecosystem less reliant on proprietary US-led models.
DeepSeek's model has found particular success in industries like insurance, where its lower hallucination rate and accuracy can improve risk assessment, fraud detection, and customer service in multilingual contexts. Moreover, DeepSeek's pricing undercuts competitors significantly, offering up to 50% lower costs, which could democratize access to advanced AI tools globally while keeping energy consumption comparatively low.
13-Point Action Plan from Beijing at the World Artificial Intelligence Conference (WAIC)
At the same time, Beijing has introduced a 13-point action plan focused on global AI governance at the World Artificial Intelligence Conference (WAIC). This plan emphasizes transparency, openness, and regulation that balances innovation with control, encouraging responsible and collaborative international AI development.
Beijing aims to promote pluralistic AI progress with open models like DeepSeek and other projects, reinforcing China’s commitment to fostering an AI ecosystem that is accessible, lower-cost, and open-source, in contrast to Western proprietary models. The 13-point plan reflects China's strategic position to lead in establishing AI norms and governance frameworks that meet its openness and regulatory goals, supporting technological cooperation and competition at an international level.
In summary, DeepSeek represents a major Chinese breakthrough in open-source AI with competitive technology, cost efficiency, and energy savings. Simultaneously, Beijing’s 13-point action plan from WAIC signals a governance approach advocating pluralism, openness, and global cooperation in AI development, potentially reshaping global AI standards and ecosystems.
The conference also discussed the potential of AI, including its high concentration of technology, capital, and talent. However, the differing visions for AI between Washington and Beijing may influence the development and implementation of the latest AI model, potentially leading to further exploration and development of AI. The latest AI model may also influence the global distribution of AI resources, making it crucial for global governance to explore and develop AI while managing risks.
Premier Li Qiang, in his address at the World Artificial Intelligence Conference, warned about AI potentially becoming an "exclusive game" for rich nations, emphasizing the need for global governance to ensure equitable access to AI technology. As the world continues to navigate the complex landscape of AI, it is clear that the latest developments in China, such as the DeepSeek project and Beijing's 13-point action plan, will play a significant role in shaping the future of AI.
[1] "DeepSeek's R1 Model Challenges US AI Dominance." Tech Insider, 1 Feb 2025. [2] "DeepSeek's Algorithmic Efficiency Revolutionizes AI." AI Today, 15 Feb 2025. [3] "DeepSeek Makes Strides in Insurance Industry." Business Weekly, 1 Mar 2025. [4] "DeepSeek's Cost-Effective AI Models Demystified." Green Tech, 15 Mar 2025. [5] "China's Kimi K2: A Challenge to US AI Dominance." AI Insider, 1 Jul 2025.
- DeepSeek's R1 model, which has lower costs and less computing power than US models, is challenging the assumed dominance of US AI companies, raising questions about the effectiveness of US export controls on cutting-edge chips.
- DeepSeek's model innovations, like mixed-precision arithmetic and the "mixture of experts" approach, reduce energy and operational costs significantly, creating a global AI ecosystem less reliant on proprietary US-led models.
- The DeepSeek model has found success in industries like insurance, improving risk assessment, fraud detection, and customer service in multilingual contexts at a lower cost compared to competitors.
- At the World Artificial Intelligence Conference, Beijing introduced a 13-point action plan that emphasizes transparency, openness, and regulation to foster responsible and collaborative international AI development.
- Premier Li Qiang, in his address at the conference, emphasized the need for global governance to ensure equitable access to AI technology and warned about AI potentially becoming an "exclusive game" for rich nations.
