2023-07-20 06:30:00 ET
Summary
- Global investors are increasingly avoiding China's AI sector due to geopolitical tensions, leading to a rise in domestic funding for the industry.
- Three former Microsoft executives, Lee Kai-Fu, Lu Qi, and Harry Shum, have become major backers of China's AI start-ups.
- Despite the shift in investment patterns, China aims to become a global AI superpower by 2030, with cumulative AI investment totaling $4 billion through June this year.
Global investors are avoiding China's artificial intelligence sector over fears of getting caught in geopolitical tensions.
Key Takeaways:
- China's AI industry is now largely funded by smaller, domestic sources, with global private equity and Western stock buyers increasingly avoiding the sensitive sector.
- Three former Microsoft executives, Lee Kai-fu, Lu Qi and Harry Shum, have emerged as major backers of some of China's most promising AI start-ups.
By Doug Young and Teri Yu
Could Chinese companies like TuSimple ( TSP ) and EHang ( EH ) soon become an endangered species on Wall Street? That outcome looks increasingly likely, as major private equity backers and stock buyers that traditionally nurtured this group of AI-related startups start to shun them over geopolitical concerns.
The shifting tide is already showing up in investment patterns for Chinese artificial intelligence ((AI)) startups, which were once highly coveted by global private equity and venture capital. When it comes to the latest round of AI developments in areas like content generation and large language models (LLMs) like those used by OpenAI's ChatGPT, China is rapidly becoming an "all in the family" affair.
That could drive names like Lingyi Wanwu , the latest brainchild of Chinese AI godfather Lee Kai-Fu, to do a U-turn from their usual preferred listing site in New York and return to either Hong Kong or the Chinese mainland to exit their investments. Even companies that have listed over the last four years, like autonomous driving specialists EHang, TuSimple and Hesai ( HSAI ), could find themselves increasingly shunned by U.S. stock buyers worried about the potential for U.S. sanctions aimed at slowing China's AI development.
As Western money keeps a growing distance from Chinese AI, the sector's list of backers is starting to look like an all-Chinese affair, with some of the most enthusiastic funders coming from the state-owned sector. A sizable chunk of funding is being channeled through a trio of Western-trained industry veterans who cut their teeth at big foreign names like Microsoft ( MSFT ) and Google ( GOOG ), before moving on to homegrown Chinese giants like Baidu ( BIDU ), and finally striking out on their own.
We'll return shortly to the stories of Lee Kai-Fu, Lu Qi and Harry Shum, whose names have become synonymous with AI in China.
Those "Three Musketeers," as we've dubbed them, all know the importance of following signals from Beijing, which has placed a huge importance on developing China's AI sector, especially around generative content and the LLMs that will power many technologies of the future. Following such winds greatly boosts the chances for success in China, opening the doors for generous government assistance, from grants, to tax breaks and funding from both state-run and private sources seeking to curry favor with Beijing.
China recently reiterated that it must seize the opportunities presented by AI, which are central to modernizing the country's industrial system. China has set an ambitious goal of becoming a global AI superpower by 2030 despite its late arrival to the game.
It definitely has some catching-up to do. Cumulative AI investment in China through June this year totaled just $4 billion, or a fraction of the U.S. total of $26.6 billion, according to market research firm Preqin. That catch-up game could become all the more difficult in the face of U.S.-led efforts to stifle China's development of industries it believes could be used for military purposes.
Such efforts have already darkened the prospects for China's aspirations to become a global leader in advanced microchips. Similar, though less visible, efforts delayed and nearly derailed the 2021 Hong Kong IPO for SenseTime ( SNTMF , 0020.HK), one of China's oldest AI companies. Other Chinese AI firms getting tripped up by sanctions that prohibit U.S. suppliers and investors from involvement with China's AI sector include speech recognition company iFlytek (002230.SZ), facial recognition firm Yitu Technology and video surveillance company Hikvision (002415.SZ), all of those China-listed.
Older names like SenseTime launched well before the wave of sanctions began, and count big Western backers like Qualcomm ( QCOM ), SoftBank ( SFTBY , 9984.T) and Temasek among their pre-IPO funders. But backers for newer names like autonomous driving AI startup Black Sesame , which last week filed for a Hong Kong IPO, read more like a who's who of lesser-known domestic names, including the likes of Dongfeng Asset Management, Hina Group, Industrial Bank and the Yangzijiang Fund.
Family affair
While China's vast economy has plenty of funds to finance its local AI sector, nothing would be possible without the necessary know-how. That's where our "Three Musketeers" of Chinese AI come in.
Lee Kai-Fu is arguably the best-known of that group, working through his Sinovation Ventures , a venture investment company founded around the time he left as head of Google's China operations, shortly before the U.S. search giant withdrew from China in 2010. Before that, he worked as a senior executive at Microsoft, which is the common denominator as a former employer of all three of our "musketeers." All three are also U.S.-educated.
Sinovation's website shows it currently has around $3 billion in assets spread across 400 portfolio companies from China's tech spectrum. Its most notable AI investments include Qingdao-based Ainnovation Technology and Lanzhou Technology.
Lee's latest AI baby, Lingyi Wanwu, is his seventh AI start-up, aiming to carve out a space in the generative AI space that burst into headlines this year with the rapid rise of OpenAI's ChatGPT, which is also tied to Microsoft. Lingyi Wanwu is aiming to build up teams focused on LLMs, AI engineering and multi-modal natural language processing. It says it has already conducted beta tests of models with tens of billions of parameters, and will soon increase its scale from 30 billion to 70 billion parameters.
Next, there's Chinese-born Lu Qi, who worked at Microsoft and Yahoo, before making headlines when he joined leading Chinese search engine Baidu in 2017 to head its AI push - including its autonomous driving initiative - as the company's president and COO. But he abruptly left Baidu just a year later. He then briefly worked for Y Combinator, but left after the U.S. tech incubator abandoned its plans to set up a China shop, and went on to open his own incubator and investment firm, MiraclePlus, in November 2019.
Last month, Lu disclosed MiraclePlus has already invested in 258 companies, including 136 AI-driven start-ups. Its latest investment was a seed funding round in May for StellarRover, which is working on an AI assistant that can better comprehend human language, according to Crunchbase.
Last but certainly not least is our third "musketeer," Harry Shum, who is a slightly different breed, lacking the same venture capital stripes as Lee and Lu. Also a native Chinese, Shum worked at Microsoft for 23 years, doing stints at its research arm and Bing program before leading its overall global strategy and broader research efforts, including AI.
Shum's most notable achievements come from his long tenure at Microsoft. One of those was developed by Microsoft's Asia AI team, its largest independent AI R&D team worldwide, called Xiaoice, whose name means "Little Ice." Xiaoice is an outgoing, Mandarin-speaking AI framework with a girly, sassy personality. Trained using data collected via Microsoft's Bing search engine, Xiaoice currently boasts 660 million global users and reaches 450 million smart devices worldwide.
Microsoft spun off Xiaoice into a separate company in 2020, naming Shum as its chairman. The newly independent company raised around $140 million in a funding round last November.
Geopolitical tensions aside, another factor keeping the big global investors away could be lack of an easy exit strategy, since many of China's AI start-ups may not be welcome on Wall Street or in other global markets due to Western prohibitions. That means that Lee, Lu and Shum may all need to find alternate listing venues to recoup their investment and the large returns they are seeking. The recent emergence of a big bumper crop of AI start-ups may also lead to worries that many may ultimately end up as duds.
With those kinds of constraints, their best options may be IPOs in Hong Kong or China's domestic A-shares market. But even those could be challenging, since many Western investors may avoid their stocks in Hong Kong, and they may have to wait for extended periods behind all the other hungry start-ups lining up to list in China's A-shares market.
For further details see:
Geopolitical Tensions Send Chill Through U.S. Listings For China AI