I was pretty disappointed by Kai Fu Li’s book AI Superpowers. He is a big expert on artificial intelligence in China but the book was mostly high-level US-China commentary and his personal life. It was surprisingly light on artificial intelligence.
In contrast, I was pleasantly surprised by Nina Xiang’s book Red AI. While not a technologist, Nina, the founder of China Money Network, really dug into what companies are doing AI in China today. And she placed this phenomenon within the bigger trend of Chinese State-directed development. This is the China AI book to read if you want details from on the ground.
Here are my 9 main take-aways from the book. Note: these are my opinions and are not necessarily the same as the author’s.
#9: State support will have a big impact on the commercialization of Chinese AI.
Nina goes into interesting detail about the long history of State monopolies in China. From the monopoly on salt during the Qing dynasty to the current control of critical national resources such as oil, electricity and telecommunications.
Kai Fu Li has a good quote about this, saying heavy handed government support “can be highly inefficient and extraordinarily effective. When the long-term upside is so monumental, overpaying the short-term can be the right thing to do.”
State support is one of the reasons why AI in China is commercializing so quickly. She also cites speed, scale, and social indifference as other drivers. In practice, watch for Chinese AI to benefit from policy preferences, talent inflows and capital support.
#8: State support will also cause problems with AI internationalization.
The current US-China problems began as complaints that Chinese State support was impacting trade with the US. The complaints ranged from unfair subsidies to outright dumping. But this quickly became a discussion about US-China technology and the relationship between companies like Huawei and the Chinese government.
I think the question at the center of this is: In a critical industry or technology, can you be both tied to the Chinese government and a multinational company? Are those two roles in conflict (whether in reality or just in perception). Huawei is at the center of this question right now.
So if the Chinese government is supporting AI development and commercialization, will this create similar problems for these companies as they go international? Can computer vision company Megvii or speech recognition company iFlyTek sell their services in the US and Europe? How about to in developing economies like Africa and Latin America? I suspect they are going to face the same questions as Huawei.
#7: Computer vision is commercializing particularly fast in China.
One of things I really like about this book is its details about many of China’s rising AI companies. And about the founders who are becoming the new generation of Chinese tech stars. And we really see this happening the most rapidly in computer vision. The book goes through some of the details of the leading companies. Here are four companies to keep an eye on.
- Founded in Shenzhen in 2014 by Xiaoou Tong of the Chinese University of Hong Kong.
- This company evolved out of the multimedia lab he had founded 13 years earlier to do research in computer vision research – including face recognition, face alignment, pedestrian detection, person parsing, and super resolution.
- Founded in Beijing in 2011 by three Tsinghua University students.
- Co-founder Yin Qi was 23 at the time – and was working at Microsoft Research Asia, where he was developing a facial recognition search engine
- Co-founder Tang Wenbin was 24 at the time – and was working on a Masters at Tsinghua with a focus on social mining and image recognition. He was also interning at Microsoft Research Asia.
- Co-founded by Yang Mu.
Note: Yang Mu and Yin Qi were both from the elite Yao Class program at Tsinghua University, which was founded in 2005 to nurture young superstars. Microsoft Research Asia is also a big source of talent for China AI.
- Founded in 2012 in Shanghai
- Co-founded by Zhu Long, who did a Phd at UCLA and a post-doc at the MIT Computer Science and Artificial Intelligence Laboratory.
- Co-founded by Lin Chenxi
- Founded in 2015 in Chongqing
- By Zhou Xi, who did a Phd at the University of Illinois.
#6: Computer vision is rapid growth tech without the losses.
These computer vision companies are getting sizeable government procurement contracts, especially from local police bureaus. You can see this in Megvii’s recent IPO filings, which show both rapid growth and nice profits. So not only does computer vision commercialize easily, it is also doing so in an immediately profitable way. That is quite a contrast to the big loss-makers we have saw in ride-sharing, retail coffee and co-working.
And yes, security and policing is a big deal in Chinese computer vision. This is about surveillance, which people argue a lot about. But it is also about dealing with a big lack of police in China overall. The book notes that China has about half the number of cops per capita as the USA. And about 1/3 as France. Computer vision is lot about addressing to this problem.
#5: After computer vision, industrial robotics is probably the next near-term winner.
I write and speak a lot about how digital China is different than digital / tech elsewhere. And one of the big ways is when digital combines with China’s massive manufacturing base. We are increasingly seeing innovation in digital and AI combining with innovation in hardware and manufacturing.
There are two sides to this phenomenon right now. The first is the B2C side where we see increasingly smart and connected devices that combine hardware and software. This includes drones (DJI), smartphones (Xiaomi, Oppo, etc.), smart TVs, smart bikes (Mobike), smart scooters (Niu) and tons of other products. These hardware plus software products are coming out of China / Asia and increasingly winning globally.
The other side is B2B where it is about giving digital upgrades to manufacturing and industry (which Jack Ma calls new manufacturing). Right now, this is mostly about industrial robots and industrial automation. The book presents some good numbers on this, showing that more industrial robots are now sold in China than in the US and Europe combined. And that sales have been growing at +50% per year recently.
In industrial robots, China has big companies, big domestic demand and State support. I expect this sector to move fast and for domestic companies to do particularly well. According to the book, the government wants Chinese companies to be 50% of the domestic market for industrial robots by 2020. And 75% by 2025. Note: it was 25% in 2017.
There are also opportunities for service robots (cleaning, reception, shopping malls, office lobbies, etc.), education robots and agriculture robots (planting, picking). But I think these are more speculative. I’m mostly watching industrial robots and industrial automation.
That’s it for the first half. In the second half, I finish my take-aways list.
I write, speak and consult about how to win (and not lose) in digital strategy and transformation.
I am the founder of TechMoat Consulting, a boutique consulting firm that helps retailers, brands, and technology companies exploit digital change to grow faster, innovate better and build digital moats. Get in touch here.
My book series Moats and Marathons is one-of-a-kind framework for building and measuring competitive advantages in digital businesses.
Note: This content (articles, podcasts, website info) is not investment advice. The information and opinions from me and any guests may be incorrect. The numbers and information may be wrong. The views expressed may no longer be relevant or accurate. Investing is risky. Do your own research.