Summary of Key Points
This article analyzes three new characteristics of contemporary technological progress: decentralization in scientific research, non-linear innovation pathways, and the unpredictability of breakthroughs. It argues that the traditional administrative-driven approach to managing technology, which resembles "growing mushrooms," is no longer suited to these new trends. The author proposes a reconfiguration of the logic for governing technological innovation—shifting from focusing on individual support to creating an ecosystem that fosters growth. Five specific measures are suggested: removing unnecessary restrictions, nurturing young talent, strengthening enterprises, improving the system, and adjusting policies to address the global competition in technology.
Detailed Explanation
1. Scientific Research Is No Longer a Privilege of Elites: Ordinary People Can Also Make a Difference
For centuries, scientific research was the domain of a few geniuses and traditional institutions (such as universities and research institutes), making it difficult for ordinary people to access even the latest knowledge. However, the internet and AI have broken down these barriers:
- Examples: Elon Musk (Tesla/SpaceX) and Huang Renxun (NVIDIA) are not from the traditional scientific community; China's DeepSeek team (developing large AI models) and Zhang Xue's motorcycle team did not rely on inherited resources but achieved groundbreaking results.
- Trend: In the future, it may become common for ordinary people to make significant contributions, and the aura of academic authority will fade as the barrier to acquiring knowledge decreases, allowing everyone to participate in innovation.
2. Innovation Is Not a Linear Process from 0 to 100: Theory Alone Is Insufficient
Previously, it was believed that science involved creating something new (from 0 to 1), while technology focused on applying that knowledge (from 1 to 100), with science considered more advanced. However, this is not the case:
- Problems: We understand the principles of lithography, but why can't we build machines like those made by ASML in the Netherlands? We know how chips work, yet our manufacturing capabilities lag behind those of TSMC.
- Reasons: Theory only covers a small part of the process; over 90% of knowledge (such as how to debug machines and solve practical problems) comes from trial and error and practical experience, which cannot be replicated by pure theory alone.
- New Trend: Science, technology, and engineering are becoming increasingly integrated. Nobel Prize winners in 2024, such as Arvind Krishna Khatriwala (AI) and David MacMillan (AlphaFold), worked on projects that combined both theoretical and practical aspects. Therefore, a evaluation system based solely on publications is no longer effective.
3. Major Breakthroughs Cannot Be Planned: Following the Lead Is No Longer Effective
In the past, we made rapid progress by copying what others did, which was controllable and cost-effective. But now, as we enter uncharted territories, planned approaches and expert votes often stifle genuine innovation:
- Examples: Huawei (5G), DJI (drones), and DeepSeek (AI) were not the result of government-planned initiatives; they succeeded through independent efforts.
- Problems: Expert votes tend to favor established, predictable directions, potentially missing out on innovative ideas that could be groundbreaking.
- Conclusion: Technological innovation has become unpredictable, requiring a complete shift in policy logic. We can no longer rely on administrative orders to guide innovation.
4. From "Growing Mushrooms" to "Creating an Ecosystem": A New Approach to Managing Technology
The traditional model involves administrative departments deciding who receives funding and awards, similar to growing mushrooms in a fixed area. This approach is no longer effective as it fails to identify cutting-edge directions or assemble the best teams. We need to shift to creating an ecosystem where policies and resources support diverse innovation:
- Removing Restrictions: Abolish unnecessary criteria (such as relying solely on publications for recognition) and decouple honors from funding and benefits, allowing academia to focus on research.
- Nurturing Young Talent: Reduce the burden of paperwork and awards for young researchers and allocate more resources to them, as they are the driving force in the fast-paced world of AI.
- Strengthening Enterprises: Enterprises are the main drivers of innovation (accounting for 79% of research investment, with private firms accounting for 60%). Policies should support both them and traditional institutions equally.
- Improving the System: Remove barriers that hinder the free flow of talent, technology, and capital, such as by sharing public scientific resources and ensuring seamless collaboration between academia, industry, and research.
- Adjusting Policies: The government should shift from being a manager to a provider of services. Support basic research with stable funding and let academia determine its direction; industrial technologies should be supported through tax incentives, government procurement, and post-project funding.
In Conclusion
Global competition now revolves around technological innovation. We must quickly reform our approach, moving from a focus on individual achievements to creating an ecosystem that promotes growth. Only by doing so can we seize the opportunities for catch-up.
(The entire article uses clear language and avoids technical jargon, making it accessible to a wide audience.)