ChinAI #92: China's Lockheed Martin, Where Art Thou?

Plus, Part 2 of our Taihe series and introducing the Ding Dare

Greetings from a land where crabapples are in bloom and where the people are deliberating over whether to walk away from Omelas

…as always, the archive of all past issues is here and please please subscribe here to support ChinAI under a Guardian/Wikipedia-style tipping model (everyone gets the same content but those who can pay support access for all AND compensation for awesome ChinAI contributors).

Edits to Browsing Taihe (a mini Defense One?) Together

Got a lot of good feedback on last week’s issue introducing Taihe. I wanted to highlight two pieces of constructive criticism from ChinAI readers:

  1. One reader made a very useful and obvious (in a good way: in the sense that once made clear, it’s obvious to all that this is right) suggestion that I should just reach out to the Taihe folks and run a Q&A by them. This checks against making these exercises too “voyeuristic.” I was able to get in touch with some of the Taihe editors this week, and have sent some questions, so hopefully we’ll be able to share some answers in next week’s post. I am going to try and take this obvious and also respectful step of reaching out to the platforms and authors we feature.

  2. Another reader took exception, with good reason, to my hot take from last week: “Here’s the reason I wanted to throw out Defense One as a possible analogue for sites like Taihe. I think if I wanted to get a good sense of what people in the U.S. military and defense industry were thinking, I would read readouts and articles in Defense One as opposed to the 300+ page reports of government defense strategies.” This reader, much more well-versed on US defense strategy than me, made a strong case for actually reading the documents (which are not 300+ pages) that guide US policy: the National Defense Strategy and the National Security Strategy. The goal here, as the reader astutely put it, is to “build off the key policy literature, creating a win-win for your research and higher-quality addition for the U.S. defense policy community!”

I wanted to use the second edit as a jumping off point for being more nuanced about how I see the value of outlets like Taihe. Again, using Defense One as an imperfect analogy, what I’m trying to get at here is that outlets like Defense One may provide some different form of value-add to learning about the US defense industry/technology base than government strategy documents. The implication being: analysts of China's defense industry may get some value out of Taihe that could supplement what they get out of PLA military academic journals or official strategy documents. Instead of framing these types of sources in a mutually exclusive way, I should have framed them as complementary and mutually beneficial.

The broader thesis I’m trying to prove with this series on Taihe, though, goes back to the original vision of ChinAI:

While traditional media and China specialists can provide important insights on [insert topic here] through on-the-ground reporting and extensive background knowledge, ChinAI takes a different approach: it bets on the proposition that for many of these issues, the people with the most knowledge and insight are Chinese people themselves who are sharing their insights in Chinese. Through translating articles and documents from government departments, think tanks, traditional media, and newer forms of “self-media,” etc., ChinAI provides a unique look into the intersection between the country that is changing the world and the technology that is changing the world.

This is a vision that builds off the work of so many other platforms, like the amazing work of China Digital Times and DigiChina. It’s also one that others are adopting as well. I’m excited about work Jordan Schneider is doing with ChinaTalk, the translation work that CSET is doing, and Politico’s new initiative (led by David Wertime) which also seems to be very much translation-based. The overall purpose is to slowly chip away at the massive language asymmetry that characterizes the current Chinese-English transmission pipeline. We’re seeing the consequences of this with how US media is “rediscovering” issues (at substantial delay) related to COVID-19 that were already covered by their Chinese counterparts — a very tangible way to represent the costs of this language asymmetry.

At the risk of being overly direct and self-referential about this: if you tell me you are an expert on China’s ___X___, you have to be able to list at least five Wechat public accounts that are focused to covering X that you follow regularly. Let’s call it the Ding Dare. If you can’t do that, I’m sorry, but you’re just not doing it right.

More specifically, this series on Taihe is making the case that we can learn a lot from newer forms of self-media and new-style think tanks even on issues one might think are relatively closed-off — like, say, the inner workings of China’s military industrial complex or, say, this week’s feature translation…

Thanks to everyone who gave feedback, and I hope to use future feedback as a way to “edit” posts.

Feature Translation: 90% of military industrial enterprises cannot introduce formal capital

Context: last week, we looked at an article from one of Taihe’s five main verticals (Transformers). This week we’re going through a 2018 article from a vertical focused on the defense industry. The article’s headline starts with a startling stat: 90% of companies in China's defense industry cannot introduce formal capital (e.g. VC funding, institutional investment as opposed to informal investors such as family or friends). Using that as a starting point and the provocative question — Can China Produce a Lockheed Martin? — as a landing point, this article provides an in-depth look at the state of civil-military integration from the anchor point of financing.

Key Takeaways — This is just going to be a list of really really interesting stats:

  • >14,000 equity investment funds in China; the ones involved in military affairs can be divided into four types: 1. central-led ones (e.g. CASIC-led VC fund); 2. local government-led CMI funds (e.g. Xi’an High-tech Investment fund); 3. M&A funds led by listed companies (e.g. Tianhai Defense and Ruiye Digital Assets have a joint M&A fund); 4. private or market-oriented institutions that also invest in defense (e.g. Fortune VC, Harvest Capital, etc.)

  • According to Zero2IPO's 2016 Top 10 VC/PE Firms in Military Industry, rough statistical estimates based on the public data regarding these top ten firms show that the total amount of capital invested in the field of military equipment in 2016 did not exceed RMB 5 billion. Far lower than hotspot fields like health, electronics, etc.

  • Why such little investment so far? One military leader who supervised equipment-related tasks told Taihe: The essence of military enterprises bringing in capital is that private capital participates in the defense industry. From a macro perspective, military enterprises are not short of money, due to the huge support from military and national defense expenditures, especially related to the core projects required by the military, so it is difficult for private capital to participate in the whole process. At the most, the military will make use of certain technologies and equipment from private companies, which means the private companies serve as an outsourcing partner. Of course, some civil-military dual-use projects with prospects for broad applications are more likely to absorb private capital to jointly develop and produce, so as to benefit from mutual gains. This is probably the key entry point for grasping civil-military integration.

  • Hence, there is a popular saying in military investment circles: "There is no venture capital in the defense industry [军工无创投]."

However — the recent push for more civil-military integration has made some progress:

  • Since 2009, 68% of private companies involved in the defense industry have come from the field of informatization; as of the end of 2017, among the units that are qualified for scientific research and production of military products, private units accounted for nearly 41%, and the pace of civilian participation in the military has accelerated significantly.

  • The State Administration for Science, Technology and Industry for National Defense’s new licensing catalogue for military equipment has been reduced by more than 60% to make it easier for private companies to navigate.

But — can a Chinese private company ever move beyond being a 2nd, 3rd, 4th-tier parts supplier, or even a builder of subsystems, to get the level where they are winning bids for large-scale complete systems such as the J-20 stealth fighter or the 055 class of destroyer. In other words, where is China’s Lockheed Martin?

  • Yu Chuanxin, Secretary General of the Civil-Military Integration Research Center of the Academy of Military Sciences, says: “I am optimistic that there will be a large number of privately-run enterprises that are globally competitive in the next 5 to 10 years. ‘China’s Lockheed Martin’ will definitely appear.”

  • “There are also some investors who believe that this is just a beautiful wish. It is a long process to obtain full trust of the military, and the possibility of properly dealing with all the sections is also very low. Of the current private military companies that are listed, strong leaders such as Wuhan Guide Infared (高德红外) have also undergone many difficulties on their road to building complete systems.”

There’s more where that come from. See FULL TRANSLATION: Outlook: 90% of companies in the military industry cannot introduce formal capital

ChinAI Links (Four to Forward)

Must-read: Global China: Technology — a special Brookings/CSET collaboration

A pretty incredible convening of the minds on the important topic of China’s growing technological reach. It’s hard to even pick out specific papers form this set, but I’d recommend Elsa Kania’s piece on “AI weapons” in Chinese military innovation, Remco Zwetsloot’s piece on China’s approach to tech talent competition, and Saif M. Khan and Carrick Flynn’s piece on maintaining China’s dependence on democracies for advanced computer chips.

Should-read: AI Governance in 2019

A year in review with observations from 50 global experts, convened by Li Hui and Brian Tse for the Shanghai Institute for Science of Science. Special shoutout to my colleagues Allan Dafoe and Markus Anderljung for their piece on the rapid growth in the field of AI governance.

Should-read: How to navigate the tradeoff between effective contract tracing and privacy

For the WashPost, GovAI researchers Toby Shevlane, Ben Garfinkel, and Allan Dafoe discuss how far we can go in reducing the trade-off between privacy and security, in the context of digital contact tracing methods. They describe the concept of “structured transparency,” which refers to the opportunity to achieve both high levels of privacy and effectiveness through the careful design of information architectures — the social and technical arrangements that determine who can see what, when and how.

Should-watch: PBS 5-hour Series on Asian Americans

Premiers May 11 and May 12 — definitely a timely and important watch. The series is going to dive in to the history of the Asian American experience, one that has to be disaggregated given the diversity that underlies this umbrella term, but also one that captures some commonalities in terms of the experience of exclusion and the status of a perpetual foreigner. H/t to my mom for sharing this with me.

Thank you for reading and engaging.

These are Jeff Ding's (sometimes) weekly translations of Chinese-language musings on AI and related topics. Jeff is a PhD candidate in International Relations at the University of Oxford and a researcher at the Center for the Governance of AI at Oxford’s Future of Humanity Institute.

Check out the archive of all past issues here & please subscribe here to support ChinAI under a Guardian/Wikipedia-style tipping model (everyone gets the same content but those who can pay for a subscription will support access for all).

Any suggestions or feedback? Let me know at or on Twitter at @jjding99