The New Startup Scene in China - and What it Means

AI rules, sector focus are defined, and opportunity is alive in the midst of geo-political uncertainty. But America remains ascendent and should lean into the best of what got it to where it is.

Painting by Yukai Du

Among the many humbling aspects of covid was how hard it was to have a reliable sense of the ground in the markets I cover. I began each day on zoom before 6:00 AM starting in Asia and ended in Latin America hours later. I spoke with entrepreneurs, other investors, policy leaders, journalists and more.

It was efficient, but misleading. It is that meal, or drink, or walk around a headquarters or neighborhood – the look in the eye and smells in the air – that tells so much and provides the trust and candor of further strings to pull on.

Little of what I wrote here on Southeast Asia and here on Saudi Arabia was clear to me in my shorts and sipping tea in my office.

I cannot remember a time in America when a unified meme around a geography has been more shared than on China. It is the only truly bi-partisan issue here in Washington, and to even question the orthodoxy begs its own risk of a version of cancellation. A great policy advisor I know with over a decade in Beijing and now located in Southeast Asia visited DC for the first time since Covid. I asked him what surprised him, and he looked over my shoulder and said: “I won’t be invited back on a panel there any time soon if I have something positive to say.”

As an investor, I’m not as interested in positive or negative only but rather the hard task of understanding what is. I am no expert on China – though in an auto-didact sense have made the attempt – impossible in many lifetimes -- to study it, surround myself with many much smarter than I, and have adored the people, history, and culture when I have visited. I do not invest in China but am very aware of its capabilities in technology and innovation and how not only that will affect America but also every rising market I do touch.

As we and China pull away from each other geopolitically, a similar distancing has happened among my sisters and brothers in Silicon Valley. The appetite of investment in Chinese tech companies and funds from the West has dried up significantly, physical interaction and travel has all but disappeared unless people are visiting families and friends. The great juggernauts like Apple, Microsoft and the hardware makers, of course, have long and historic connection but many are famously exploring manufacturing partnerships as widely spread as India, Vietnam and Mexico.

In increased isolation, or at least distance, even this community has gravitated around its own memes. They all fall under the umbrella of two themes. The first is centrally planned innovation simply cannot work over time – real innovation comes from hundreds and thousands of entrepreneurs testing, competing, failing – and this weakness will only compound in the new eras of AI, bio tech, quantum computing and more. The second is China is shooting itself in the foot attacking the very enormous and success tech enterprises and entrepreneurs built in the days of relative greater openness. Look at Jack Ma, or the failed Ant IPO, or the wiping out of the entire ed-tech space – and many examples since of government and regulatory intervention -- and the map appears very clear.

When I was in Southeast Asia a few weeks ago, a successful entrepreneur who had built several enterprises in China looked at me puzzled when I suggested the best days of China’s startup scene may be behind them. He shrugged and said, “There is no time I’d rather be investing in Chinese startups then now. The approach will be different, the government will show a strong hand when they are concerned about corruption or risk to stability. But there are enormous serious challenges to be fixed. The next era is less about another delivery app, but rather revolutions in health care, electric vehicles, bio tech, climate change and more.” It struck me that these sectors will be difficult, at least for venture capital from abroad, to have easy exits as they seem more intertwined with government than consumer internet that yielded much in returns to date. Another investor was more circumspect: “It will be much harder than ten years ago for sure – but key sectors will absolutely yield significant returns.”

Notwithstanding collectively this feedback, to say the least, jarring from the conventional wisdom back here at home. So, I decided to unpack it.

Because I cannot go to China for a while, I called on some of the smartest people I know long engaged with the ecosystem there. It is a testament to our times that they wanted to speak only “off the record.” But their reactions were staggeringly consistent, and I share them without editorial comment. I also rely on Tech Buzz China as a good independent resource on companies and data there.

Chinese startups are as fixated on Generative AI as in the States. All the engineering talent is aiming here, AI has been a leading major in university for some years, and while early and there are many examples of experimentation in chatbots, automating ecommerce and the back end, the speed of adoption and depth is light speed. Despite the Great Fire Wall, no one has trouble accessing OpenAI and the others from America. As this chart shows, a year dated but indicative, China is both ahead and behind the United States overall:

They are not close to us on LLM, and catching up could take even years, but are having significant success in areas like computer vision with extensive visual data available due to all the cameras, and a regulatory environment that both controls data but allows data collection where the West puts more limits.

The next generation of entrepreneurs are hyper focused on the “big problems” and the way to think about the hand of government is to understand the importance of social and political priorities over economic. Many in the West glossed over this in the earlier hyper-growth, less-supervised days of tech in China, but it was true then also. The Ant IPO action was not some random crack down, but a genuine concern of insufficient regulatory management in areas that affect the pocketbooks of millions. “It is crucial to know the lanes, which are becoming increasingly clearer, and then entrepreneurs will be unleashed,” one investor told me. The story in China is no longer about fintech, consumer internet, retail and ecommerce, or ed tech. Look for significant innovation in healthcare, EV, advanced manufacturing and climate tech. “There is less low hanging fruit than a decade ago, but there is a LOT of fruit” one advisor told me.

The talent is better than ever. At one level, there is little new here – China has had superb tech talent for some time. What is different is that the ever-increasing number of new computer software engineers is led be an entire generation that has learned crucial lessons building tech companies that have reached massive scale. What Silicon Valley long called “the PayPal effect” – hundreds of entrepreneurs who learned how to build there spun off to start their own enterprise – in Beijing is called the “ByteDance effect.” Interestingly, however, and as indicative of the economic realities there now, all I spoke with also noted that this increased talent is having trouble finding jobs.

The companies rising and leveraging newest technologies like AI are better than ever. The examples are legion, but several people told me, speaking of ByteDance, that it is one of the best examples they know of an “AI-first thinking” company. Data moves almost every decision, constantly improving existing product performance and innovation. Their recent launch of Lemon8, going to toe to toe with Instagram in America and beyond, will use data for ever improving relevance to encourage sharing and purchasing in food, beauty, travel and wellness. I’ll be focusing more on EV generally; autonomous vehicles like WeRide; robotics delivery solutions like Keenon; image processing such as sensetime spanning health care, autonomous vehicles, augmented reality and more and HikVision in Internet of Things (IOT) and security solutions; and WuXi Biologics in contract r&d for a wide spectrum of biologics. Each of these small samplings have global operations and sales, including often in America.

Much analysis both in business and geopolitics focus on what America misunderstands about China, and my sources say there are many in the tech fields. Americans still often underestimate the speed of product shipping and business model transition in China which gives them greater flexibility in product development and release. For all the talk of there not being a large cost differential any more in China, the best engineers still often cost three to five times less on average there. One noted that America too often thinks of “first” as the same thing as “leading.” “Copycat” became a pejorative word in America, synonymous with low quality knockoffs. But in fact, the greatest enterprises in China were learners who often adopt the best technology and solutions to their significant market and constantly innovate and iterate on top of that. And Americans, who cherish efficiency and productivity sometimes underestimate the sheer will Chinese enterprises have to throw people at any problem at almost any cost if they see the long-term profit potential.

But the Chinese misunderstand a good deal about American startups also. The United States is a melting pot of needs, not as homogeneous a market as China can be, requiring both more nuanced product and marketing skills and embracing talent that comes from every corner of the world. They often assume markets outside of Asia value low cost first and foremost, but that is often a mistake when high reliability and consistency is valued especially in enterprise software. They underestimate the power of American brands globally including in China.

Perhaps most missed of all among Chinese tech watchers on America is the belief that America and the West is in decline. There are no certainties in life, but there is a reason why the United States has set the rule book in almost any major breakthrough technology since World War II and playing “catchup” in a world of AI may not even be possible in the uncertain but exciting paths that AI are opening. I also picked up on not inconsequential tech talent in China aspiring to leave and build outside of there.

Certainly, and hard to predict, is what Chinese regulation may do with all this. Translated last week was an early draft of AI regulations – emphasis on draft. But much of the language sounds tailor-made to slow innovation. Language like: "Organizations or individuals that use generative AI to provide services such as chat, text, image, or audio generation … including providing programmable interfaces … bear responsibility as the producer of the content generated by the product." Or "When generated content that does not conform to the requirements of these Measures is discovered during operations or reported by users … repeat generation is to be prevented through such methods as optimization training within three months."

Though the first cut of language from the EU had some similarity.

America may be well ahead overall but, at the same time, America should be cautious in assuming that these leads last. The lesson of the Web2 era may be less that the world copied what the West began. But, instead, that China and many nations took learning and leap-frogged over existing infrastructural challenges, innovated not only in and for their markets but by global standards, and have never had a more prepared and experienced and larger talent bases for what is to come.

A leading figure in Singapore told me – and I’ve heard similar language in a recent trip around Europe – “The best way for America to compete is, well, to compete. Offer the most compelling product, at the most compelling price, lean into your innovation and trusted global partnerships and get your own house in order is a good place to start.”

One friend from the American tech world offered me a compelling five-point plan to compete: 1) lean into the multi-decade partnership between the US Government and private sector, pushing to rapidly develop new technologies from the private sector, rapid adoption by the government, and harnessing the fruits of a true free market system; 2) rebuild the severely degraded relationships between the US government and the private sector, especially in the tech sector, declaring this a top strategic priority; 3) remove any regulatory constraints to this end and unleash innovation across sectors; 4) ensure that all steps are taken to unleash specifically startup competition, including US startups challenging US incumbents and to avoid entrenching permanent oligopoly of big US tech companies; 5) ensure that US companies are on a level playing field competing with China and others anywhere abroad.

I haven't heard anything so blunt, clear and concise out of Washington thus far. In fact, much action and language appears to discourage or blunt our advantage.

Two approaches to innovation and startups are being tested in historic times with historic consequences. Where we go from here begins with calmly understanding where we are right now.