2023.06.10 Views:
At the recent Shenzhen Qianhai " The Venture Capital Day Of Shenzhen" forum, Zhang Wei, Chairman of CoStone Capital，expressed optimism about the AI productivity revolution brought by ChatGPT and the emergence of new application scenarios in industries such as office, education, healthcare, and imaging. Based on it,CoStone Capital will focus on investing in companies that specialize in computing power, storage, data, and large-scale models in the future.
Zhang Wei stated that artificial intelligence has gone through three major stages of development from 2012 to the present.
The first stage, from 2012 to 2016, saw limited applications of artificial intelligence in fields such as facial recognition and license plate recognition, represented by companies like Google, META, Baidu, and Alibaba. Zhang Wei mentioned that during this period, CoStone Capital invested in companies like SenseTime, CloudWalk Technology, and Innovation Works.
The second stage, from 2017 to 2021, witnessed the expansion of AI applications into fields such as internet entertainment, smartphones, smart healthcare, and autonomous driving. In this stage, CoStone Capital invested in companies like 4Paradigm and SmartMore, according to Zhang Wei.
The third stage began in 2022, where AI disruptively revolutionizes productivity, and the birth of ChatGPT4 has the potential to reshape all industries and create new applications, represented by companies like OpenAI.
Regarding ChatGPT, Zhang Wei mentioned that its core lies in algorithms, computing power, and data, which require a mighty infrastructure for artificial intelligence. For instance, OpenAI's technological breakthrough relies on powerful Nvidia GPU chips. "There is a significant gap between China's current GPUs and those of Nvidia and AMD. Additionally, there is a noticeable disparity in storage chip technology compared to foreign companies like Samsung."
Looking ahead, Zhang Wei recommended that China should focus on investing in and supporting infrastructure companies related to algorithms, computing power, data, and more. He specifically emphasized the significance of the "three factors: proactive government, innovative capital, outstanding entrepreneurs" in areas like Shenzhen, Beijing, and Shanghai, where the government is aiming to seize the next round of AI industry leadership.
Firstly, the government should continuously upgrade and establish new-generation computing centers to meet the requirements of the new AI industry. At the same time, existing supercomputing centers should be studied to adapt to the needs of the new generation AI industry.
Secondly, the government needs to support the development of local GPU and storage chip companies, covering areas from computing power to storage.
Furthermore, the government should establish data sharing platforms.
Additionally, the government should strongly support the development of innovative capital and utilize market-based approaches to promote industry competition, ultimately selecting outstanding AI entrepreneurs.
According to Zhang Wei, the United States has constructed the Silicon Valley innovation ecosystem, which involves collaboration between companies, universities, national laboratories, independent research labs, platform service providers, public and private clients, multiple institutional entities, startup companies, and investors.
According to Zhang Wei's introduction, the United States has established the Silicon Valley innovation system in the development of the AI industry. This system involves collaborations between companies and universities, national laboratories, independent research labs and universities, platform service providers and public or private clients, multiple institutional entities and various partners, as well as cooperation between large corporations, institutions, startups, and investors.
Drawing on overseas models, Zhang Wei also mentioned that Shenzhen needs to formulate a comprehensive industrial plan by planning production lines and designing system layouts. This entails creating a blueprint to establish an overall industry planning framework. Simultaneously, Shenzhen should focus on talent introduction, identifying and recognizing entrepreneurs, and supporting the development of private enterprises. "More importantly, Shenzhen needs to invest substantial resources to strongly support the development of the AI and chip industries.
Over the past year, the bull market returned, with technology leading the way.
As I said back in October 2024, the rally that began on September 24 was not merely a rebound, but a reversal. Since then, China's capital markets have undergone a broad revaluation, moving from undervaluation toward fair value.
The Shanghai Composite Index reached its highest level in a decade. The "Yi-Zhong-Tian" trio—Eoptolink, Zhongji Innolight and TFC Communication, three prominent Chinese optical communications stocks—saw their share prices multiply severalfold in a single year. Cambricon's market capitalization briefly exceeded RMB 600 billion. Elsewhere in the technology sector, triple-digit price-to-earnings ratios and valuations exceeding RMB 10 billion became commonplace.
Some investors were greedy; others, fearful. Still others asked me: Are technology stocks in a valuation bubble?
My answer is that what we are witnessing is the hunger and anxiety of an era, reflected in the capital markets. Once you understand this, the current fervor for technology investing is no longer difficult to explain.
Investment has nothing to do with macroeconomics, but everything to do with geopolitics.
China-US relations will shape the investment landscape for decades to come. Geopolitics has played a major role in shaping the hunger and anxiety of our times, fundamentally reshaping the way we invest.
What, then, lies at the heart of China's hunger and anxiety?
The Fourth Industrial Revolution is already underway, yet China has not fully overcome the technological shortcomings left by the Third. There is no time to lose.
This is why some people wonder: Why do Chinese computing chip companies command such high valuations when their technologies still lag so far behind NVIDIA's? Why does NVIDIA trade at just 50 times earnings, while a company hailed as "China's answer to NVIDIA" commands a multiple of 300?
To me, this gets the logic exactly backward.
It is precisely because these companies cannot catch up with NVIDIA in the near term that they command a higher strategic premium. The wider the technological gap, the greater the strategic value.
The further something lies beyond our reach, the stronger our hunger. The harder it is to catch up, the deeper our anxiety. If China were to achieve a complete breakthrough in computing chips one day, valuations might actually come down.
Where can one find China's most spectacular natural landscapes?
Much of the answer, I believe, lies in and around the Hengduan Mountains: the Meili Snow Mountains, Jade Dragon Snow Mountain, Mount Gongga, Mount Siguniang, Jiuzhaigou, and many others. Vast differences in elevation and deep fault lines have given rise to landscapes of extraordinary grandeur.
Hard-tech investing follows a similar logic.
Technological gaps, discontinuities in capabilities, and missing links in industrial systems: a leap in difficulty brings a leap in valuation, and a successful breakthrough can, in turn, bring a leap in returns.
This logic extends well beyond computing chips. It is equally evident in memory semiconductors, commercial spaceflight, and other sectors where China still faces substantial technological gaps with other countries.
It was precisely this conviction that led CoStone Capital, starting in 2016, to decisively focus its investments on three areas: hard tech, emerging industries, and life sciences and healthcare. In this context, "hard tech" refers specifically to important industries in which China faces a technological gap with the United States.
Thanks to this forward-looking strategy, CoStone Capital established an early presence across the full value chains of key sectors, including semiconductors, artificial intelligence, and robotics.
Looking ahead, the frontiers of technological progress and industrial upgrading are boundless. Nor can the gaps in hard technology be closed overnight. The twin windows of opportunity presented by technological innovation and the development of homegrown alternatives to foreign technologies have only just begun to open.
The road ahead is long, but the future holds immense promise.
As an investment firm, CoStone Capital will continue to invest where the hunger and anxiety of each era run deepest, move forward with the tensions that shape our times, and extend our reach toward the most challenging—and most important—frontiers.
Spring 2026