Summary of Key Points
The main conclusion from Kevin Kelly's (KK) discussion with Chinese scientists is that the real challenge to China's technological innovation is not artificial intelligence (AI), but a lack of originality. The conversation covered topics such as research systems, cultural perspectives, talent development, diversity, and the development of AI. It was pointed out that while China is highly efficient at scaling innovations from 1 to N (from existing technologies to new applications), it faces bottlenecks in generating truly groundbreaking ideas from scratch (from 0 to 1). The root causes include a research mechanism driven by key performance indicators (KPIs), a cultural tradition of obedience to authority, a lack of interdisciplinary exploration, and insufficient diversity in innovation.
1. Research Systems: KPIs Limiting Originality
The biggest problem with China's research system is the use of KPIs as a guiding principle. Professor Lu Bai mentioned that universities and research institutions focus on achieving publications in top journals like CNS (Cell), Nature, and Science, as well as obtaining funding and promotions. As a result, scientists spend their energy on completing tasks rather than engaging in curiosity-driven exploration. For example, to get promoted to professor, one must publish a certain number of high-impact papers or secure significant funding—similar to students memorizing exam questions without time for meaningful inquiry.
Original research is often risky and time-consuming; Darwin spent 20 years studying evolution, and Einstein developed the theory of relativity not for awards. However, the current system does not allow scientists to devote five years to a single question without worrying about funding, as KPIs require immediate results. This environment discourages risk-taking, which in turn reduces innovation.
2. Cultural Barriers: Obedience Hinders Questioning
China's long-standing culture values obedience and respect for authority, which can stifle innovative thinking. Professor Lu Bai noted that Western scientists are more willing to challenge authorities and engage in debate, whereas Chinese scientists tend to be submissive. For instance, students and young researchers are hesitant to question their mentors, even though scientific progress often requires overturning established theories (like Copernicus challenging geocentric models or Newton questioning Aristotle's mechanics).
KK also emphasized the importance of embracing failure. Scientific journals only publish successful outcomes, but failures contain valuable lessons. Edison failed thousands of times with the light bulb, learning from each attempt. He suggested that we should celebrate failures by asking scientists about their setbacks and what they have learned from them.
3. Gathering Talents: Creating Environments for Creative Competition
KK discussed the concept of a “Scenius” (a gathering of talented individuals who inspire and compete with each other), such as the Cavendish Laboratory (which produced 29 Nobel laureates) or Bell Labs (which invented the transistor). To foster such environments, it’s essential to have a playful and serious approach to research. While many Chinese scientists are highly motivated, there is a lack of those who pursue research for its own sake. For example, Claude Shannon, the founder of information theory, created a seemingly useless device to study control mechanisms. KK suggested that even if only 1% of a team consists of such “players,” they can significantly boost creativity.
4. Diversity: Both Human and Artificial Intelligence
Diversity is crucial for innovation. Professor Wang Liming pointed out that Chinese innovation tends to follow similar trends, with many researchers focusing on the same topics. KK believes that diversity can break this pattern:
- Human Diversity: Immigration brings diverse perspectives, which is beneficial for innovation (e.g., Google Street View was inspired by artists using cameras in carts). China could consider attracting international immigrants, though language barriers may be overcome with translation technology.
- AI Diversity: AI should be designed to think in unique ways, not just imitate humans. For example, AI should be rewarded for using novel methods to solve problems, even if they differ from human approaches.
5. The Future of AI: Practicality Over Cost
Regarding the future of AI, KK predicts a phase of cost competition, where Chinese models may become more affordable (10 times cheaper than European or American ones). While this can be an advantage, Professor Lu Bai emphasizes that extreme performance is still necessary for solving complex scientific issues. KK also stressed the importance of user experience; people often choose expensive models because of their usability, not just their capabilities. Chinese AI should focus on making it easy for ordinary users to integrate technology into their daily lives.
6. Final Thoughts
To overcome innovation barriers, China needs to change its research system to focus less on KPIs and more on fostering a culture that encourages playful and innovative thinking. It should also promote diversity among researchers and AI systems. Innovation is about doing things that no one has done before, even if they seem useless at first glance—these are often the breakthroughs of the future.
(The entire analysis is presented in plain language to make it accessible to non-financial professionals.)