Analysis: The US and China race for AI hegemonyーNHK WORLD-JAPAN NEWS

Key Concepts

• AI Dominance Race: The competition between the United States and China for leadership in artificial intelligence development and implementation.
• Private Sector Investment: Financial capital injected into AI companies by non-governmental entities.
• AI Models: Sophisticated algorithms and systems designed to perform tasks that typically require human intelligence.
• Open Source: Software or models whose source code is made publicly available, allowing for modification and distribution.
• Enterprise AI Adoption: The integration of AI technologies and solutions by businesses.
• AI Strategy: Government-led plans and initiatives to foster AI development and adoption.
• AI Cold War: A term suggesting an escalating geopolitical tension and competition centered around artificial intelligence.

US vs. China: The AI Power Struggle

The United States and China are engaged in a significant power struggle for dominance in the rapidly evolving field of artificial intelligence (AI). This competition extends beyond trade disputes and export controls, focusing on technological advancement and market leadership.

Private Sector Investment in AI

A key indicator of AI development is private sector investment. The transcript highlights a striking disparity:

• US Companies: Have invested over 11 times more in AI than China.
• China: Is actively working to catch up, emphasizing technology development and rapid social implementation.
• Japan: Trails significantly, with investments being 117 times less than the US.

Leading AI Models and Players

The landscape of AI models is dynamic, with leading positions shifting:

• Past Leader: OpenAI's ChatGPT was previously considered the top AI model.
• Current Leader: Google's most advanced AI model now ranks at the top of the Global Intelligence Index.
• Emerging Chinese Player: Moonshot AI, a Chinese startup backed by Alibaba, has developed an open-source model that ranks fifth, just behind Google, Anthropic, and OpenAI. This model was developed at a fraction of the cost of its US rivals, who spent billions.
• Geographic Concentration: Most of the top AI players are based in either the US or China.

End User Adoption and Market Trends

The adoption of AI models by end users reveals distinct market preferences:

• Japanese Companies: Many companies in Tokyo primarily use US AI models.
  • Mitsubishi UFJ Financial Group: Has partnered with OpenAI, aiming to become an "AI native" organization.
  • OpenAI: Identifies Japan as its second-largest market for enterprise AI adoption after the US, with a commitment to providing "safe and responsible AI services and products to bring sustainable growth to Japanese society and industries."
  • Anthropic: OpenAI's rival, opened its first Asian office in Tokyo, focusing on business partnerships and creating economic value with Japanese businesses.
• Chinese Models Gaining Traction: Despite the preference for US models, Chinese AI models are starting to gain traction in Tokyo.
  • Alibaba's QWEN model: Is mentioned as an example of a Chinese model gaining interest.

Japanese Performance in the AI Race

Japan is experiencing rapid growth in its startup ecosystem, despite historically being a challenging environment:

• Sakana AI: A standout example, recently raised 20 billion yen in funding from both Japanese and overseas investors (including Spain's Santandere Group).
  • Valuation: This funding pushed its valuation to 400 billion yen (approximately $2.6 billion), reportedly the largest ever for a startup in Japan.

Government Approaches to AI

Governments are taking distinct approaches to AI development and integration:

• United States:
  • AI Action Plan (July): Structured around three pillars:
    1. Accelerate AI innovation.
    2. Build American AI infrastructure.
    3. Lead in international AI diplomacy and security.
  • Stated Goal: "to do whatever it takes to lead the world in artificial intelligence."
• China:
  • AI Plus Initiative (August): A three-phase roadmap aiming for deep integration of AI into society and the economy by 2035.
• Japan:
  • Acknowledged Lag: The government admits it is lagging behind in AI adoption.
    • Company AI Usage Rate: 55.2%
    • Individual AI Usage Rate: 26.7%
    • These rates are significantly lower compared to the US and China.
  • Planned Basic AI Strategy: The government intends to announce a strategy by the end of the month to address low adoption and limited investment relative to its economic scale.

The Billion Dollar Question: Who Will Win?

The outcome of the AI race remains uncertain, with differing perspectives from industry leaders:

• Nvidia CEO Jensen Huang: Stated, "China is going to win the AI race."
• Former Google CEO Eric Schmidt: Argued, "The vast majority of governments and countries will end up standardizing on Chinese models, not because they're better, but because they're free." This highlights the potential impact of cost as a deciding factor.

Challenges and Future Outlook

• For Japan: The challenge lies in balancing the appeal of cost and convenience with the necessity of safeguarding data protection and national security.
• For US and China: Tensions are expected to escalate, with some experts terming the situation an "AI cold war."

Conclusion

The AI race is a complex geopolitical and technological contest primarily between the US and China, with significant implications for global economic and security landscapes. While the US currently leads in investment and has strong private sector players, China is rapidly advancing with a focus on implementation and cost-effective solutions. Japan faces the challenge of increasing its AI adoption and investment while navigating the strategic interests of the two AI giants. The role of cost and accessibility of AI models, particularly open-source options, could be a critical determinant in the long-term outcome.