In a more recent note, Louis-Vincent Gave suggests China’s recent launch of Deepseek challenges prevailing beliefs underpinning the bull market in U.S. growth stocks.
Another Sputnik Moment
Gavekal
Louis-Vincent Gave
Research
5 Pages
Key Takeaways
DeepSeek's Efficiency: DeepSeek was developed using Nvidia's H800 chips—designed to comply with U.S. export controls—at a cost of approximately $5.6 million, significantly less than the over $100 million reportedly spent on training models like ChatGPT-4.
Challenging U.S. Tech Dominance: China's advancements in AI, exemplified by DeepSeek, question the prevailing belief that the U.S. holds an insurmountable lead in cutting-edge technology.
Market Implications: The emergence of cost-effective AI models like DeepSeek may impact the valuations of U.S. tech companies heavily invested in AI, as the competitive landscape evolves.
Strategic Considerations: Gave suggests that the U.S. may need to reassess its approach to technological competition, considering the rapid advancements and resource investments made by China in AI and related fields.