Anna's Deep Dives

_{Just facts, you think for yourself}

A. Broken Supply Chains: New Tech Groups and a More Split Digital World

The U.S.-China chip fight is breaking up global chip supply chains. This was very clear on April 17, 2025. Nvidia said it would lose $5.5 billion in its first business quarter of 2026. New U.S. export rules on its H20 AI chip caused this. These rules could cut Nvidia's sales in China from about 20% to almost nothing. AMD might also lose up to $800 million.

Over 100 Chinese groups now have strict U.S. export rules against them. Even with these problems, the global chip market is expected to grow. It should reach $697 billion in 2025, up from $627 billion in 2024. But the industry has a big shortage of skilled workers. It needs over 100,000 new skilled people every year. Forecasts show a shortage of 1 million such workers by 2030.

Countries are forming 'tech groups' because of this. One important group includes the U.S., Japan, and the Netherlands. This group aims to work together on controlling chip tech and production equipment. This teamwork is important when you look at global R&D spending. In 2022, China spent $811 billion on research and development. This was almost 88% of the $923 billion the U.S. spent that year.

The U.S. CHIPS and Science Act gives about $52 billion to $53 billion to help local chip making. A national goal for the U.S. is to make 20% of the world's most advanced chips by 2030. New chip-making centers are also appearing, like in Panama. The World Fab Forecast said global chip-making ability would grow by 6.4% in 2024. This shows that production is spreading out geographically.

The world's economy breaking into pieces greatly affects trade and supply chains. By 2023, 74% of EU-based global companies said they planned to change their supply chain plans. This was a big jump from 42% in 2018. Experts warn this breakup might cause global GDP to fall, maybe between 0.2% and 7.0%. Companies are busy moving production to friendly countries ('friend-shoring'). This is to depend less on China. The U.S. CHIPS Act alone has led to announcements for over 80 new chip-related projects in America.

B. What's Next for Innovation: A Boost or a Bust from Competition?

The tough global competition for the best AI chips in 2025 makes the future of innovation complicated. It could start a new age of tech. Or, it might turn into a race where quality drops if countries only care about their own interests instead of global progress. U.S. export rules, which started in early 2025, affect 140 countries, mostly aiming at China. But eighteen friendly countries get mostly free access to U.S. AI chips.

Big U.S. companies like Nvidia could see up to 50% of their AI chip sales affected. Nvidia makes about 56% of its money from other countries. Its stock fell about 2% after these rules were announced. Meanwhile, Chinese startup DeepSeek released competing AI models in early 2025. These models perform well and often cost less to develop. This suggests U.S. rules might accidentally speed up some kinds of innovation in China.

Countries like India are also improving their AI skills. India's national AI plan has a budget of about $1.3 billion (₹10,371.92 crore). But India has trouble getting top workers and enough money for research, falling behind U.S. and Chinese spending. Wider trade fights, like recent U.S. taxes on Chinese solar cells, could make it even harder for scientists from different countries to work together.

The global AI chip market itself is growing fast. It's expected to hit $50 billion in 2025. Nvidia has about 25% of the market, Intel has 15%, and AMD has 10%. The fight for new ideas (intellectual property) is still intense. Over 1,000 AI-related patents were filed around the world in early 2025 alone. Open-source AI models, like some from DeepSeek, are becoming more popular. Programs like the Model Context Protocol (MCP) try to help AI models connect better worldwide. But international rules for AI are still very mixed. While countries like India and France are announcing big new AI funding, high costs to follow rules and for training are problems, especially for smaller companies.

C. Tech "Islands": A Risk to Global Progress?

Having different AI rules around the world creates a clear risk. It could lead to "tech islands" forming. New ideas might grow well in separate places, but overall global progress could be slowed down. The European Union's strict tech rules are an example. Actions like Apple's €13 billion tax bill and Google's €2.4 billion fine show the EU's strong position. Some critics say these rules can slow down quick innovation.

If these different rules become more common, the EU and other areas risk becoming these 'islands.' In that case, only countries with looser rules or those pushing their own national tech standards might do well. This could limit new tech from spreading globally and being developed together. The U.S. officially supports a 'balanced approach' to AI. It wants to manage risks while helping growth. But its own export rules naturally lead to a different kind of tech separation. If countries create tech standards that don't work together or greatly limit data sharing between countries because of security worries, global science progress could slow. This could mean people do the same research twice and human progress slows down.

D. Separate Work, Surprise Discoveries?

While breaking into pieces has risks, tough international competition can also be a strong push. It can lead to different groups working on new ideas separately. This, in turn, might lead to surprise discoveries from different tech environments. China's strong effort to become self-sufficient in tech shows this. Facing U.S. rules, Chinese companies like Huawei are speeding up development of their own AI products. For example, Huawei's Ascend 910C AI chips were planned for large-scale production by 2025. These chips are meant to replace the powerful foreign chips they can't easily get.

China's AI industry is now worth over $70 billion. This system includes about 4,300 companies. It has gained from about $912 billion in general tech investment over the last ten years. This focused work is creating a special AI innovation environment. The appearance of companies like DeepSeek shows this possibility. DeepSeek's success with powerful AI models, made for much less money than in the West, goes against common ideas. It shows that different ways of doing things can lead to surprising and game-changing results. As different countries follow their own tech goals, based on their own needs or ethics, they might find new uses for AI. These different efforts could, strangely enough, make the global tech world richer, even if there's less working together at first.

E. Power Shifts: How AI Chips Are Changing Global Politics

The global fight to be the best in AI is greatly changing international power. The U.S. and China are currently leading this game-changing race. Both countries see AI leadership as very important for their national security and future economy. They are investing billions to improve military skills and secure tech advantages.

China wants to be the world leader in AI by 2030. The U.S. defense budget, about $717 billion, strongly supports AI development. This competition pushes smaller countries to pick sides based on their tech needs. Chips are basic to AI, often called the 'new oil.' Japan and the Netherlands are important suppliers of manufacturing equipment. Taiwan and South Korea are very good at making chips. The U.S. has limited China's access to these advanced chips.

China, on the other hand, controls about 70% of the world's rare earth minerals. These minerals are vital for tech, including making chips. The U.S. still depends on China for these resources. Chinese company DeepSeek released an AI model in 2025 for only $6 million. This model reportedly did better than U.S. competing systems that cost over $100 million. This low cost makes U.S. tech companies rethink their ideas.

Trade numbers show these tensions. ASEAN's trade with China hit $234 billion in early 2025. Its trade with the U.S. was $476.8 billion in 2024. The U.S. has put taxes on some Chinese imports as high as 145%. Taiwan's TSMC is still very important for both the U.S. and China. The U.S. military runs over 685 AI-related projects. Ukraine's use of AI for watching and for self-driving vehicles shows AI's growing role in the military. The global AI market is expected to grow from about $600 billion in 2025 to $1.81 trillion by 2030. Being the leader in AI chips will greatly shape global politics in the 21st century.

Subscribe to Premium to read the rest.