GenAI and the Japanese Economy: the Digital Deficit and Economic Security

Naoyuki Fukuda at the Japan Research Institute has recently published a paper titled "Generative AI and the Japanese Economy: Reducing the Digital Deficit and Ensuring Economic Security".

Fukuda begins by introducing the transformative potential of generative AI, based on deep learning. Generative AI is described as a revolutionary tool capable of creating content and revolutionizing how individuals live and work. It is identified as a primary driver of economic growth and social development, potentially reshaping the global competitive landscape. The United States, led by companies like OpenAI and Google, and China, with Baidu and Alibaba, are at the forefront of this technological advancement, leading to a continued struggle for technological supremacy.

The paper emphasizes that while generative AI offers immense possibilities, Japan faces challenges related to its digital deficit (the imbalance of payments for digital services) and economic security. It is suggested that over-reliance on foreign IT services and cloud computing resources increases the digital deficit and poses economic security risks.

Impact of Generative AI on the Japanese Economy

Fukuda argues that generative AI holds the potential to revolutionize various industries in Japan, including manufacturing and healthcare, leading to improved productivity and competitiveness. In manufacturing, generative AI can enhance design and development processes, optimize part shapes, and improve quality control. In healthcare, it can accelerate drug discovery, reduce the burden on medical professionals, and improve diagnostic accuracy.

However, Japan's reliance on overseas cloud computing infrastructure, intensified by the demands of generative AI, threatens to further inflate its existing digital deficit. The digital deficit is predicted to reach approximately 6 trillion yen in 2024 and is forecast to further increase up to 10 trillion yen, offsetting benefits from other sectors.

Fukuda stresses the importance of developing domestic AI capabilities to mitigate risks associated with over-reliance on foreign providers. Economic security is threatened when AI and cloud resources are dependent on foreign nations. Service interruptions and data leaks could have serious socio-economic consequences.

AI Policies and Investments in Different Countries

The paper compares AI policies and investments in key regions:

• United States: The US pursues a two-pronged strategy led by the private sector with robust national security support. The CHIPS and Science Act allocates significant funding to bolster semiconductor production.
• China: China adopts a government-led digital economic strategy with substantial investment in AI research and development.
• European Union: The EU prioritizes digital sovereignty and regulatory leadership, with stringent regulations like GDPR and the AI Act, while also fostering domestic AI capabilities through initiatives like GAIA-X.
• Japan: Japan aims to strengthen its generative AI development capabilities through initiatives like GENIAC. However, private sector investment in Japan lags significantly behind the US and China.

Fukuda identifies that there is a huge difference in investment. In 2023, private investment was about 10 trillion yen in the US and 1.2 trillion yen in China, while Japan was only 100 billion yen. Japan ranks 12th in the world, which is a significant difference compared to the US and China. This investment lag leads to delays in domestic LLM development and computing resource development.

Addressing Overseas Dependency

Fukuda suggests strategies for Japan to mitigate its dependence on overseas AI resources and enhance its competitive edge:

• Expanding Private Investment and Implementing Domestic AI: This strategy includes encouraging private sector investment in domestic AI development, focusing on enterprise implementation and government support.
• Exploring Domestic AI for Industrial Specialization: This approach investigates the potential of industry-specific LLMs developed domestically.
• Developing Cloud Computing Infrastructure: This entails improving and constructing cloud infrastructure that is critical to generative AI.
• Fostering and Recruiting Talent: This means to invest in AI talent to develop and strengthen the AI workforce.

Ultimately, the goal is to transition Japan into an AI exporting nation.

Moving Towards AI Export

Fukuda proposes strategies for Japan to evolve beyond AI consumption to becoming an AI exporter, positing that Japan's AI strategy should transcend merely mitigating its digital deficit and aspire to a more ambitious objective: transitioning into an AI exporting powerhouse, ultimately achieving a surplus in the digital balance. This shift requires a proactive and multifaceted approach, leveraging Japan's unique strengths while addressing existing weaknesses.

Leveraging Existing Strengths and Cultivating Niche Specializations

Japan already possesses significant strengths in specific sectors, particularly manufacturing and robotics. To become a major AI exporter, it needs to integrate AI solutions into those sectors and enhance them.

• Robotics and Automation: Japan can capitalize on its expertise in robotics by developing AI-powered robotics solutions for manufacturing, logistics, healthcare, and other industries. This involves creating AI algorithms that optimize robot performance, enhance automation capabilities, and enable robots to perform more complex tasks.
• Healthcare and Medical Technology: Japan has a well-established healthcare system and a strong tradition of medical innovation. It can become a leader in AI-powered healthcare solutions, such as AI-assisted diagnostics, drug discovery, and personalized medicine. This involves developing AI algorithms that analyze medical images, predict disease outbreaks, and identify the most effective treatments for individual patients.

Developing Specialized AI Models Tailored to Specific Industries

Rather than attempting to compete directly with the US and China in the development of general-purpose AI models, Japan should focus on developing specialized AI models tailored to the specific needs of its key industries.

• Manufacturing AI: Developing AI models for optimizing manufacturing processes, predicting equipment failures, and improving product quality.
• Robotics AI: Developing AI models for enhancing robot navigation, object recognition, and human-robot interaction.
• Healthcare AI: Developing AI models for diagnosing diseases, analyzing medical images, and personalizing treatments.

Fostering International Research Collaboration and Knowledge Sharing

Establishing strong relationships with international research institutions and AI communities is essential. Collaborating on research projects, sharing knowledge, and participating in international AI forums will help Japan stay abreast of the latest advancements in the field.

Promoting Data Accessibility and Sharing

Creating a data ecosystem that allows researchers and developers to access high-quality data is crucial for AI innovation. This involves establishing clear data governance frameworks, promoting data sharing initiatives, and investing in data infrastructure.

Standardization and Certification

Japan can contribute to global standardization efforts in AI, promoting ethical and responsible AI development. Developing certification programs for AI products and services can enhance trust and confidence in Japanese AI technologies.

Conclusion

Fukuda concludes that while generative AI provides opportunities for Japan to enhance productivity and competitiveness, it also presents challenges related to the digital deficit and economic security. Collaborative efforts between the government and private sector are crucial for fostering domestic AI capabilities, mitigating risks, and ultimately transforming Japan into an AI exporting nation.