Those who view China merely as the ‘world’s factory’ are still looking at a country that no longer exists. In recent decades, the Asian giant has become a continent-scale laboratory capable of developing proprietary chips, training foundational AI models, creating vertical digital ecosystems, and deploying applications for hundreds of millions of people in a matter of weeks. It’s more than technology: it’s culture, strategy, and execution.
I was able to observe all this up close during an immersive on-site experience with companies like Huawei, Alibaba Cloud, Meituan, Kwai, SenseTime, and Nio, as well as innovation centers in Beijing, Hangzhou, and Shanghai. I also participated in the 8th World Artificial Intelligence Conference (WAIC), which brought together global leaders around the theme ‘Global Solidarity in the AI Era.’ The fieldwork allowed me to see how technology, culture, and strategy intertwine to create nationwide impact.
The Chinese machinery starts long before the first prototype. Culture and education are at its core. In a country that was never colonized and carries over 5,000 years of history, trust relationships are built slowly, but execution, once decided, is swift. Work follows an intense rhythm (the famous 9/9/6 model), and education is treated as a strategic vector for innovation, with pressure and investment to develop talent on a massive scale.
This cultural foundation meets a business and governmental ecosystem that operates in a coordinated manner. Huawei, for example, allocates 20% of its revenue to R&D and develops its own AI models; Alibaba Cloud vertically integrated its entire tech stack and created the Qwen model family; Meituan handles 150 million daily orders by combining multiple services in a super app; and Kwai already connects over 60 million users in Brazil to social commerce, a phenomenon that accounts for over 25% of e-commerce in China. Models like the X27 (shopping transformed into a mega live commerce studio) and vehicles like Nio’s, featuring robotically removable batteries in 3 minutes (BaaS system, battery as a service) and integrated virtual assistants, illustrate how innovation permeates entire sectors.
What’s impressive is not just what China creates, but the speed and scale at which it implements. AI models trained for specific sectors go live quickly, and autonomous agents are already present in retail, healthcare, mobility, and public management—all supported by a data infrastructure and digital penetration reaching over 99% of the population.
Brazil, on the other hand, is advancing in a more fragmented way. We have technical talent, creativity, and a significant domestic market, but face structural barriers: slower regulatory frameworks, still-modest R&D investments, and little integration between government, businesses, and academia. Our digitalization is progressing but lacks the same technological vertical integration and a robust national strategy that aligns sectors and defines long-term priorities.
Of course, the Chinese model isn’t simply replicable. It’s deeply rooted in its history, political system, and culture. But there are clear lessons: invest heavily and continuously in research; treat technology as a sovereignty asset; create mechanisms for companies to innovate not just in products but in infrastructure and standards; and, above all, coordinate efforts, recognizing that digital competitiveness requires a decades-long vision, not election cycles.
The world is heading into an era where AI, data integration, and applied innovation will define not just markets but also each nation’s place on the geopolitical map. China has already understood this and is executing. Brazil has the foundation to learn fast and apply ambitiously. How do we implement, with coordination and speed, what’s already proven to gain global competitiveness?
*Gustavo Pinto is a senior researcher at Zup Labs, a division dedicated to research and development (R&D) in Generative AI, where he leads applied research for Zup—a technology company within the Itaú Unibanco group—and its clients. With a Ph.D. in Computer Science from UFPE, Gustavo has authored over 100 scientific papers in software engineering.
