Expert Unveils U.S.-China AI Divide Using Grok 3: "Grok 3 reveals China-U.S. AI divide"
Grok 3 highlights the ongoing divide in artificial intelligence between China and the United States, according to an expert analysis.
TroibNews 
The launch of Elon Musk's Grok 3 AI chatbot has sparked renewed conversations about the global AI race, but according to Tian Feng, former dean of SenseTime’s Intelligence Industry Research Institute, its true significance lies in the differing developmental strategies employed by China and the U.S.
The U.S. continues to advance AI primarily through substantial funding and powerful computing capabilities, while China adopts a more cost-effective, efficiency-oriented method, as demonstrated by DeepSeek.
In a phone interview with CN, Tian noted that U.S. companies, supported by significant capital and advanced chips, are propelling AI advancements through computational prowess. This strategy, illustrated by Grok 3, emphasizes innovation at considerable costs. With its financial supremacy, the U.S. tends to prioritize maintaining a technological edge over China without as much focus on cost-effectiveness.
Conversely, China emphasizes achieving maximum efficiency with available resources, as seen in DeepSeek’s highly optimized training and inference processes. The training cost of DeepSeek V3 is approximately one-tenth that of OpenAI’s GPT-4, and Tian anticipates that AI computing costs will decrease even further in the next year. This efficiency-focused approach enhances accessibility to AI technology, expediting its adoption as a practical tool in various industries.
The role of open-source in AI's future is also pivotal, according to Tian. Unlike Grok 3, which isn't open-source, DeepSeek openly shares its findings and allows for commercial use without licensing fees. This transparency encourages participation from global developers, creating a "flywheel effect" that accelerates iterations, commercialization, and scientific advancements.
Tian asserts that for China to cultivate a robust AI ecosystem, it must foster a strong AI community through open collaboration. Open-source initiatives will enable Chinese researchers and developers to unite, promoting larger-scale innovation.
In addition to AI models, Tian identifies China's manufacturing sector as a key area for AI integration. He highlights how Tesla's gigafactories provide Elon Musk with a wealth of real-world data, enhancing his AI development efforts.
However, Tian argues that China's industrial diversity could be as much as 100 times that of the U.S. and encompasses a larger scale.
The integration of AI into manufacturing is poised to be transformative for China. As Tian emphasizes, "No scale, no intelligence." The maturation of AI, particularly as it moves toward Artificial General Intelligence, relies on extensive data and widespread adoption. China's vast industrial base positions it strongly to lead advancements in AI-driven manufacturing.
Looking ahead, Tian envisions two distinct trajectories for AI development:
1. High-cost, high-performance models – The U.S. will continue to heavily invest in premium AI models for advanced applications, leveraging its financial resources to maintain a competitive edge.
2. Affordable, mass-scale AI – China will concentrate on enhancing AI access and efficiency, striving to maximize performance with existing hardware while minimizing costs.
As the AI race accelerates, DeepSeek has already become a standard against which major U.S. and Chinese models are evaluated. Tian believes that DeepSeek’s efficiency and cost benefits will keep it competitive in the shifting AI landscape.
Ultimately, the future of AI will be influenced by these divergent strategies—one fueled by unlimited resources, the other by a focus on efficiency. The key question remains: which approach will achieve greater sustainability in the long term?
Ramin Sohrabi for TROIB News
The U.S. continues to advance AI primarily through substantial funding and powerful computing capabilities, while China adopts a more cost-effective, efficiency-oriented method, as demonstrated by DeepSeek.
In a phone interview with CN, Tian noted that U.S. companies, supported by significant capital and advanced chips, are propelling AI advancements through computational prowess. This strategy, illustrated by Grok 3, emphasizes innovation at considerable costs. With its financial supremacy, the U.S. tends to prioritize maintaining a technological edge over China without as much focus on cost-effectiveness.
Conversely, China emphasizes achieving maximum efficiency with available resources, as seen in DeepSeek’s highly optimized training and inference processes. The training cost of DeepSeek V3 is approximately one-tenth that of OpenAI’s GPT-4, and Tian anticipates that AI computing costs will decrease even further in the next year. This efficiency-focused approach enhances accessibility to AI technology, expediting its adoption as a practical tool in various industries.
The role of open-source in AI's future is also pivotal, according to Tian. Unlike Grok 3, which isn't open-source, DeepSeek openly shares its findings and allows for commercial use without licensing fees. This transparency encourages participation from global developers, creating a "flywheel effect" that accelerates iterations, commercialization, and scientific advancements.
Tian asserts that for China to cultivate a robust AI ecosystem, it must foster a strong AI community through open collaboration. Open-source initiatives will enable Chinese researchers and developers to unite, promoting larger-scale innovation.
In addition to AI models, Tian identifies China's manufacturing sector as a key area for AI integration. He highlights how Tesla's gigafactories provide Elon Musk with a wealth of real-world data, enhancing his AI development efforts.
However, Tian argues that China's industrial diversity could be as much as 100 times that of the U.S. and encompasses a larger scale.
The integration of AI into manufacturing is poised to be transformative for China. As Tian emphasizes, "No scale, no intelligence." The maturation of AI, particularly as it moves toward Artificial General Intelligence, relies on extensive data and widespread adoption. China's vast industrial base positions it strongly to lead advancements in AI-driven manufacturing.
Looking ahead, Tian envisions two distinct trajectories for AI development:
1. High-cost, high-performance models – The U.S. will continue to heavily invest in premium AI models for advanced applications, leveraging its financial resources to maintain a competitive edge.
2. Affordable, mass-scale AI – China will concentrate on enhancing AI access and efficiency, striving to maximize performance with existing hardware while minimizing costs.
As the AI race accelerates, DeepSeek has already become a standard against which major U.S. and Chinese models are evaluated. Tian believes that DeepSeek’s efficiency and cost benefits will keep it competitive in the shifting AI landscape.
Ultimately, the future of AI will be influenced by these divergent strategies—one fueled by unlimited resources, the other by a focus on efficiency. The key question remains: which approach will achieve greater sustainability in the long term?
Ramin Sohrabi for TROIB News
