Greetings from a world where…

fall is in full bloom

…As always, the searchable archive of all past issues is here. 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).

Around the Horn (3rd edition)

You know the drill by now:

• short preview of 10 articles related to ChinAI — all published this past week and sourced from scans of WeChat accounts and groups

• reply to the email and/or comment on the Substack post with the number of the article you’re most intrigued by, and we’ll translate a couple for next week!

• some added weight to votes from subscribers that are supporting ChinAI financially

1) One-minute survey: What are the science and tech ethics issues that you are most concerned about?

Summary: Background info on a survey of Chinese academic circles re: their views on the most pressing science and tech ethics issues. Article summarizes the effort and links to the full text of the questionnaire.

Source: Duan Weiwen, a professor in the philosophy of science and technology (Chinese Academy of Social Sciences), who is an active researcher of AI ethics.

2) Sex, Love, and Robots

Summary: A letter to the reader on the state of sex robots, written from the perspective of a robot from 2050 named Avary.

Source: 造就 (Zaojiu) — very interesting Shanghai-based platform that started out doing events similar to Ted Talks, but now is doing a range of creative media

3) Beijing’s enthusiasm for self-driving cars — even Wall Street has felt it

Summary: Open to anyone in Beijing, the most popular “tourist attraction” in Beijing is a trial of Baidu Apollo’s self-driving cars (only on designated test routes).

Source: 量子位 (Qbit AI) — AI-focused news site that pumps out many articles about AI on a daily basis. This one’s a lengthier report.

4) DJI gets back on the road

Summary: Wang Tao (AKA Frank Wang), the founder and CEO of DJI, is reflecting on new directions for DJI, the biggest seller of consumer drones. What challenges has DJI encountered, and where does its road lead?

Source: 雷锋网 (Leiphone) — a long-time source for ChinAI translations, which I think of as China’s MIT Tech Review. See ChinAI #60 for a deep dive into this source.

5) How do SMEs acquire digital intelligentization solutions? Data to help understand these transactions

Summary: Drawing from research on nearly 1000 small, medium, and micro businesses, Synced’s think tank has published a new report on digital intelligentization in China.

Source: 机器之心 (Synced) — very similar to Qbit and Leiphone, and also has an impressive longform portal (机器之能/jiqizhineng)

6) Li Mu, Liu Qun, Liu Yang, Zhu Jingbo, Zhang Min: Current bottlenecks in machine translation

Summary: Readout from the China Conference on Machine Translation, held Oct 10-12, where there was a forum on current bottlenecks of machine translation, featuring some big names, including Liu Qun — chief scientist of speech semantics at Huawei’s Noah’s Ark Laboratory.

Source: AI科技评论(aitechtalk) — focuses on in-depth reports on developments in the AI industry and academia.

7) Chinese chip projects welcome an “unfinished tide,” six 10-billion-level projects suspended

Summary: Summary of an investigative report by Xinhua's Outlook (Liaowang) Magazine on six major semiconductor projects, which have stalled. Reporters visit semiconductor projects and find empty offices overgrown with weeds.

Source: 瞭望 (Liaowang) Am curious to dig more into this Outlook Magazine report, which is written for elite government officials, and posts unusually candid reports from time to time.

8) When PaddleHub meets WeChat Mini-Programs

Summary: Apparently there’s been some excitement by developers over Baidu’s PaddleHub (an extension of its PaddlePaddle framework). This OSChina contributor tries to use PaddleHub’s poetry writing and art style transfer modules to build a WeChat mini-program.

Source: OSChina — portal that covers China’s open source community

9) Be wary of “Connecting Everything”: The Autistic Symptoms of Facial Recognition Technology

Summary: a somewhat long-winded attack on the intrusiveness of facial recognition by Yu Shengfeng, a law professor at Beihang University. Includes a tenuous comparison between the privacy-corrosive effects of the technology and the symptoms of autism.

Source: 南都公益基金会 (Narada Foundation) — ranks as one of the top five Chinese foundations in terms of charitable activities; this is a longform article published by its Narada Insights platform

10) Breakthrough! Six banks fined more than 40 million RMB for infringing personal information | DataLaws

Summary: On the same day that a draft law on personal information protection was announced, many banks were fined for infringement of the personal information of consumers.

Source: 数据法盟 (Datalaws) — a non-profit academic platform that focuses on data privacy and data security

ChinAI Links (Four to Forward)

Must-read: Thousands of Weibo accounts have been deleted as China’s government cracks down on free speech

Shen Lu’s latest story, for Rest of the World, about how a Chinese government crackdown on speech resulted in the deletion of thousands of accounts on Weibo, the country's last major platform for free expression. Zhahao, or “account bombing,” where pro-nationalists report dissenters and get them censored, have become incredibly frequent. As a result, some users have become more cautious with their conversations, while others have grown used to losing accounts and starting over, or learned where to buy burner accounts (which they quickly run through).

Should-read: What’s Going Wrong with Chinese Literature in Translation

By Dylan Levi King, for RADII, the list of best Chinese fiction rated by users on Douban looks very different from the Chinese literature that makes its way into English translation. "What makes it into English translation is often shaped by the idea that Chinese fiction’s main function is to explain China, and by two sides wrangling over what story Chinese literature should tell," Dylan writes. The bigger issue, as he notes, is that there’s so little Chinese literature that makes it into English at all.

Should-read: Chinese Perspectives on AI and Future Military Capabilities

Late to read this impressive Aug 2020 report by Ryan Fedasiuk. Instead of focusing on high-ranking military leaders’ statements and official PLA policy documents, he analyzes 58 journal articles written from 2016–2020 by officers, defense industry engineers, and academics involved in the day-to-day development and deployment of AI. One interesting key finding is: “Chinese experts tend to overestimate U.S. military AI capabilities, relative to open-source reporting.”

Should-watch: GovAI webinar series featuring Margaret Roberts on Censorship’s Implications for Artificial Intelligence

I’m really excited to be a discussant in this webinar next Wednesday, October 28th. GovAI is hosting Margaret Roberts, a professor at UC San Diego and leading expert on China’s censorship regime. She’ll be presenting work co-authored with Eddie Yang.

The topic: how censorship has affected the development of Wikipedia corpuses, which are in turn regularly used as training data that provide inputs to NLP algorithms. They show that word embeddings trained on the regularly censored Baidu Baike have very different associations between adjectives and a range of concepts about democracy, freedom, collective action, equality, and people and historical events in China than its uncensored counterpart Chinese language Wikipedia. They examine the origins of these discrepancies using surveys from mainland China and their implications by examining their use in downstream AI applications.

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 chinainewsletter@gmail.com or on Twitter at @jjding99

Greetings from a world where…

the term “best and brightest” is often misused, failing to carry the tone or irony that the original intended.

…As always, the searchable archive of all past issues is here. 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).

Feature Translation: Tsinghua/Peking University all go to the U.S.? Five years of data analysis is here!

Context: From sciencenet.cn (科学网), platform for science-related news and blogs, which in the past has featured posts with spicy debates and scandals involving Chinese scientists. This article is more matter-of-fact, drawing on data recently released by Tsinghua and Peking re: the whereabouts of their graduates. The article’s hook takes on a common narrative on the Chinese interwebs — that Tsinghua University and Peking University have become a base for training talents for the U.S.

Key Takeaways:

• Over the past five years, about 16% of Tsinghua and Peking graduates have gone abroad for further studies. This has held steady, and the article states that it is not as high as some Internet rumors. Figure below displays the Tsinghua talent trajectories. The second row gives the graduates that go abroad for further studies; the three columns are divided up into Bachelor’s graduates, Master’s graduates, and PhD graduates. For those interested in diving deeper, I translated the figures and put up the tables for both schools in this Google Sheet.

• Still, the U.S. is overwhelmingly the top choice for Tsinghua/Peking graduates who do go abroad: 70%! of them pick the U.S., with the UK at a distant second (8%). Graph below shows this for Peking University. The first horizontal bar shows that 61% of Peking graduates, among those who study abroad, pick the U.S. as their destination.

• They tend to go to pretty good schools, at least according to the Times Higher Education World University Rankings. In the past five years, more than 60% of Tsinghua University graduates studying abroad have studied in the top 50 universities in the world, compared to 57% of Peking University graduates.

• Huawei has become the company that recruits the most graduates from Tsinghua and Peking, with a total of 1,248 people in the past five years. Here’s a translated version of a table from the article, showing the top 5 employment destinations over the past five years:

More detailed breakdowns on those three trends in the FULL TRANSLATION.

ChinAI Links (Four to Forward)

Must-read: Estimating the Number of Chinese STEM Students in the United States

Speaking of getting hard data to dispel Internet rumors…

Jacob Feldgoise and Remco Zwetsloot in a CSET analysis: “In recent years, concern has grown about the risks of Chinese nationals studying science, technology, engineering and mathematics (STEM) subjects at U.S. universities. This data brief estimates the number of Chinese students in the United States in detail, according to their fields of study and degree level. Among its findings: Chinese nationals comprise 16 percent of all graduate STEM students and 2 percent of undergraduate STEM students, lower proportions than were previously suggested in U.S. government reports.”

Should-read: Robert Lighthizer Blew Up 60 Years of Trade Policy. Nobody Knows What Happens Next

By Lydia DePillis for ProPublica, an excellent dissection of Trump’s trade policy, which doesn’t grade out well. Here’s a piercing comment from a former USTR staffer — “There’s a cadre of 75-year-old white men in the trade realm who just want to turn back the hands of time,” said another staffer. ‘They don’t understand that the world has changed.’” But, overall, a balanced take that I think is pretty fair to Lighthizer. Before reading, I didn’t know that he had proposed getting rid of investor-state dispute settlement, for instance. A lot of good details about China trade policy too.

Should-read: Facial recognition data leaks are rampant in China as Covid-19 pushes wider use of the technology

Xinmei Shen of SCMP covers facial recognition systems in residential neighborhoods. Article has some comments from me as well as some of Prof. Lao Dongyan’s pushback against residential neighborhood adoption of these systems — recall that we highlighted this story in an Around the Horn issue back in September.

Should-listen: Intertwined relations: China, the US, and the global trade of AI

I’m the guest on this latest episode of The Economist’s Asia Perspectives podcast. It’s a quick-hitter, 30-min episode — thanks to Jason Wincuinas and their podcast team for the good, wide-ranging conversation.

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 chinainewsletter@gmail.com or on Twitter at @jjding99

Greetings from a world where…

another white paper roams free

…As always, the searchable archive of all past issues is here. 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).

Feature Translation: White Paper on AI Governance Summary Slides

Context: From two entities we’ve featured quite a bunch in past issues: CAICT and the AI Industry Alliance. Had some serious difficultly actually getting the full pdf of the White Paper, which was just published in September — if anybody else has more luck, let me know. The article announcing the white paper directs you to follow CAICT’s WeChat account and get the pdf link from there: http://www.caict.ac.cn/kxyj/qwfb/bps/202009/P020200928368250504705.pdf.

Unfortunately, I had a real tough time downloading the full white paper using this link. The original article link did have some summary slides, though, so I converted into Google slides and added some selective translations as comments. All slides that I translated should have a comment icon with a 1 in the middle.

Key Takeaways:

• Slide 5 highlights 3 technical characteristics of AI that make it more difficult to govern than other technological domains: 1. general-purpose nature (leads to risks that are more widespread); 2. dependency on data (leads to results that are more uncontrollable); 3. algorithmic blackbox (generates processes that are more difficult to explain).

• Slide 6 emphasizes the risk of AI influencing the political process and smearing (抹黑) political figures. This was one of three risks highlighted at the societal level. The other two were: intensification of unemployment/wealth gap, and encroachment upon videos of incident (侵害实践频发), which I take to refer to the manipulation of security videos?

• A lot of priority on self-regulation by companies, which is not surprising with AIIA as a co-drafter. See slides 10 and 22 (which identifies companies as the main governance entity in the near-term stage of AI governance)

Perceptions of other countries’ approaches to AI governance (slide 14, 19):

• The paper frames the U.S. as using ethical norms to guarantee national security. They recognize the U.S. as having “published the world's first AI ethics principles for military uses, grasping hold of the ‘power over explanation’(解释权) for (military AI) regulations.” Germany’s goals for “AI Made in Germany” is mentioned. China’s approach is described as advocating for the development of responsible AI

• Paper also analyzes countries’ efforts in subdomain-specific governance. In autonomous vehicles, it evaluates the U.S. has having a very clear autonomous vehicle strategy. South Korea is judged to be the first to put forward autonomous driving safety standards. Other subdomains covered are deepfakes, smart finance, and smart medicine.

EXCERPTED TRANSLATION: SUMMARY SLIDES ON AI GOVERNANCE WHITE PAPER

ChinAI Links (Four to Forward)

Should-read: Patent Landscape for Computer Vision: United States and China

CSET data brief by Simon Rodriguez, Autumn Toney, and Melissa Flag drills down into the subdomain of computer vision and finds that China has overtaken the U.S. in patent filings in computer vision (based on data from 1790 analytics).

Note: I think we should be pretty skeptical to draw too many conclusions from patent data, especially since this brief makes no mention of patent quality. Studies have shown that China’s patent stats are inflated by both universities and companies taking advantage of patent subsidies to produce large quantities of low-quality patents. Only 4 percent of patent applications filed in China are then filed in other jurisdictions, which is a key marker of quality. The comparable figure for the U.S. is 32 percent. I cited these points in past testimony to U.S.-China Economic and Security Review Commission, in which I concluded: “China is not poised to overtake the U.S. in the technology domain of AI.”

That being said, the approach to focus on one domain of AI is one we could all learn from. The authors capture this well, “Research and reporting on this topic tend to generalize AI, yet treating it as a singular entity with a homogenous development landscape loses sight of variability in both research and potential applications. In order to effectively compare AI production between countries, it is necessary to drill down into the subdomains of AI and identify exactly where and how nations truly lead.”

Should-watch: Bridging AI's Proof-of-Concept to Production Gap

Andrew Ng, founder & CEO of Landing AI and founder of deeplearning.ai, discusses key challenges facing AI deployments and possible solutions, ranging from techniques for working with small data to improving algorithms' robustness and generalizability to systematically planning out the full cycle of machine learning projects.

Should-read: The Future of Military Applications of Artificial Intelligence: A Role for Confidence-Building Measures?

In Orbis, Michael C. Horowitz, Lauren Kahn, and Casey Mahoney explore confidence-building measures as a form of information-sharing and transparency-enhancing arrangements to enhance strategic stability. Their aim is to “speed the learning process about the implications of military applications of AI in ways that reduce the risk that states’ uncertainty about changes in military technology undermine international security and stability.”

Should-read: Techie Software Soldier Spy

Longread by Sharon Weinberger on Palantir, as it prepares to IPO. Exposes some creation myths, and gets into the nitty-gritty of how tech is actually deployed in the military. Concludes with this: “So why are people still so excited about Palantir? One former national-security official told me the company is now famous for being famous, sort of like the Kardashians.”

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 chinainewsletter@gmail.com or on Twitter at @jjding99

Greetings from a world where…

众人拾柴火焰高 [When everyone adds fuel, the flame burns brighter] zhong4ren2 shi2 chai2 huo3yan4 gao1

…As always, the searchable archive of all past issues is here. 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).

Feature Translation: Tencent and Manufacturing?

Thanks everyone for taking part in the second round of ChinAI Around the Horn. Another close vote among 1, 4, 3, and 9. I chose #1 — “Tencent, a Manufacturing Re-evaluation” — since it had the most support among the choices that paying subscribers voted for (and tied for most votes overall).

Context: in-depth article on Tencent’s effort to transform China’s manufacturing sector from 机器之能/jiqizhineng (Synced), which has been the source for many previous ChinAI issues on smart manufacturing and the industrial Internet: #58, #70, #78.

Key Takeaways:

• Wu Xiaobo, a financial writer, predicted in 2017 that in the next few years, 80% of small and medium-sized enterprises in traditional manufacturing will go bankrupt

• These companies have two choices: 1) transform via informatization (think robotic arms and industrial Internet); 2) implement the “servicification” of manufacturing — for instance, if you’re a construction machinery manufacturer, you also have to provide maintenance services for the equipment after the initial sale.

• Tencent is working at multiple entry points in the manufacturing chain, including: marketing, data collection and monitoring of equipment, industrial vision in production lines, and ecosystem partners in independent software vendors.

Mini case studies of all 4 vectors:

• Marketing: WeChat Enterprise/WeChat Work (similar to Slack) helps LingLong Tire target hundreds of millions of users directly, rather than going through 400 dealers and 50 or 60 automobile factories. Linglong was ranked first among Chinese companies in the tire original equipment list, though Michelin and four other foreign companies placed abote it (insert bad headline about the race for tire dominance).

• ICT monitoring and data collection on equipment: SANY Heavy Industry can complete the monitoring of 400,000 engineering equipment, reaching early warning of equipment failures 6.5 hours in advance, and the early warning accuracy rate is 87%.

• Industrial vision opportunities in production: Tencent Cloud’s work with China Star Optoelectronics —  the first domestic AI recognition project for LCD panel defect types, ADC (Auto defect Classification, has been repeatedly brought up as a typical example of industrial Internet intelligent manufacturing. 

• Ecosystem partnerships with independent software vendors: "The service provider does something similar to a bricklayer, building the raw materials provided by Tencent into a house where businesses can live." Zhu Ning, the founder of Youzan, once described the relationship between the WeChat ecosystem and Youzan.

Article posits some intriguing differences between Germany and China re: their approach to manufacturing transformation:

• In Germany, top companies within the manufacturing industry led the charge; in China, the existence of major consumer Internet companies dictates that they will play an important role in the transformation of manufacturing.

• Germany more focused on optimizing production processes, whereas Chinese companies see more opportunities in improving the range of manufacturing services —to me, this one was more of an unproven generalization but an interesting theory

• Longtime readers are probably tired of me harping on this point, but read the full translation and you’ll find that there are zero references to the U.S. in this entire article. In contrast, references to Japan and Germany abound. AI is not a two-player game.

A lot more details in the FULL TRANSLATION: Tencent, A Manufacturing Re-evaluation

Middleware — JeffJots

Back when Bill Simmons used to write articles, instead of spouting bad anti-Lakers takes on podcasts, he used to end his mailbag columns with an outrageous-crazy-funny email from a reader followed by: “Yup…these are my readers.”

After last week’s issue, in which I suggested “middle platform” as a translation for 中台, many readers corrected me that “middleware” was the better fit. Thanks especially to Noah and Yorwba for their insights. Middleware is software that connects the frontend and backend systems via APIs and the like. Alibaba did not invent the concept, as I implied, but they might have popularized the concept in China. Yup…these are my readers.

I had encountered the term a couple times before in the course of dissertation research, and this discussion sparked me to look back through some notes, so let’s get into another edition of JeffJots:

• From a Casper and Whitley 2004 Research Policy article: They find that Sweden is extremely successful in the “middleware” sector, which differs from more traditional software that comes in ready-for-consumer-use programs. Instead, middleware technologies help link basic architectures to standard application software. For typical middleware products, think: software that transforms the content of web servers into formats suitable for mobile phones, or secure payment systems used in e-commerce applications.

• Why is Sweden so successful in this sector? Why is the UK not as successful, even though it is very strong at innovating in standard software? Per Casper and Whitely’s explanation it’s because Sweden’s firms (led by Ericsson) developed and shared common technical standards, which is especially important in a sector like middle ware, which requires significant coordination. They tell a similar story about Germany’s success in the life sciences, in which Germany dominates the “middleware” of biotech (platform biotechnology) but is less successful in developing new therapeutic drugs.

• This article is cited in Jingjing Huo’s book How Nations Innovate (p. 73), in which Huo argues that a nation will be successful at a particular type of innovation based on the nation’s “variety of capitalism.” Coordinated market economies, like Germany and Sweden, will be better at process innovation, as the middleware case shows. Liberal market economies, like the U.S. and the UK, will be stronger in product innovation.

What does this all mean? There’s a lot to unpack here and a lot of contention over “varieties of capitalism,” but here’s one clear takeaway: I often get asked in events about whether U.S.-China tech competition is a zero-sum game. What the above examples of specialization, like Sweden in middleware, remind us of is: Trade 101. Yes, of course there are strategic aspects of trade and there will be sensitive areas where competition will be of a “you-win, I-lose” nature, but these are the exceptions, not the rule. The baseline and default is positive-sum.

ChinAI Links (Four to Forward)

Must-read: How My Mother and I Became Chinese Propaganda

It’s impossible to capture what this story is about. The New Yorker’s short description — “Immigrant struggles in America forged a bond that became even tighter after my mother’s A.L.S. diagnosis. Then, as COVID-19 threatened, Chinese nationalists began calling us traitors to our country” — doesn’t do it justice. But I guess I’d say it’s a story of survival — in the face of diseases of all forms: A.L.S. and COVID, racism, and Chinese propaganda.

Listen to the author, Jiayang Fan, talk about this piece on the Longform podcast, which also highlights some of her other great writing.

Should-read: What’s Behind Technology Hype

Jeffrey Funk for Issues in Science and Technology pierces though the hype: “After years of hype about AI, some traditionally optimistic voices are finally beginning to temper their exuberance. A March 2019 article in IEEE Spectrum argued that Watson, IBM’s AI division, had “overpromised and underdelivered” on personalized health care applications, and shortly thereafter IBM pulled Watson from drug discovery. An April 2019 article in Technology Review went further with a title “This Is Why AI Has Yet to Reshape Most Businesses.” My forthcoming article in IEEE Spectrum (“Why Projections for AI’s Economic Benefits Are Overly Optimistic”) demonstrates that the most-well-funded AI start-ups are not targeting productivity-enhancing applications, and many are likely incurring huge losses.”

Should-read: State of AI 2020 Report

For the third year, Ian Hogart and Nathan Benaich review the state of AI. Had a chance to review the report before they published it, and they have a lot of good stuff on China’s AI ecosystem as well.

Should-read: New Chinese Ambitions for 'Strategic Emerging Industries,' Translated

On September 8, 2020, the National Development and Reform Commission, Ministry of Science and Technology, Ministry of Industry and Information Technology, and Ministry of Finance published this document: 'Guiding Opinions on Expanding Investment in Strategic Emerging Industries and Cultivating Strengthened New Growth Points and Growth Poles.” It’s been jointly translated by DigiChina and CSET: Elsa Kania, Ngor Luong, Caroline Meinhardt, Ben Murphy, Dahlia Peterson, Helen Toner, Graham Webster, and Emily Weinstein, and edited by Ben Murphy and Graham Webster.

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 chinainewsletter@gmail.com or on Twitter at @jjding99

Greetings from a world where…

Jeffrey Ding @jjding99

In 2017 Japanese firms, labs, etc. filed more invention patents in China (40,908) than U.S. firms, labs, etc (36,980). Same trend holds for those that were granted. Question: when's the last time you read an article about Japanese firms in China? Source: China Stats Y'book 2018.

September 25th 2020

10 Likes

Eugene Han, a PhD fellow at RAND who researches China tech, commented, “The last time was reading about a Japanese government initiative to fund shifting production out of China. Otherwise almost nothing over the past year(s).”

Indeed the major story about Japanese companies in China, at least this year, was that Japan was subsidizing 87 companies to shift production lines out of China. First reported by Nikkei Asian Review, this story was later picked up by the New York Times and the Washington Post. It makes sense why this is the only story that made it past the editor’s desk: probably got good clicks, tied to a specific policy, and fits with the DECOUPLING! narrative.

What’s missing is an understanding of the bigger picture, which the stats in my tweet point toward. Here’s the context that these articles lack: there are 33,050 Japanese companies in China, and the proportion of Japanese companies that chose to shrink, relocate, or withdraw from the Chinese market was the lowest it’s been in the past five years — 4.3 percent points less than the ratio in 2015. That’s from a recent 2020 Japan External Trade Organization (JETRO) White Paper on “Japanese Companies and the Chinese Economy.” According to JETRO’s survey, 93.8% of Japanese companies in China said they would either maintain the status quo or expand their business scale in China. Via Sina Finance’s (Chinese-language) readout.

I think this is a really good example of a blindspot in English-language coverage of China. As David Wertime’s latest Politico China Watcher reminds us: Earth to Washington and Beijing: It’s not all about you.

…As always, the searchable archive of all past issues is here. 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).

Around the Horn Again

Let’s run it back. Same structure as ChinAI #109:

• short preview of 10 articles related to ChinAI — all published this past week and sourced from scans of WeChat accounts and groups

• reply to the email and/or comment on the Substack post with the number of the article you’re most intrigued by, and we’ll translate a couple for next week!

• some added weight to votes from subscribers that are supporting ChinAI financially

My hope is that folks researching this space and related areas can get more familiar with these types of sources, so I put more emphasis on linking to past ChinAI coverage of specific platforms. Okay, let’s go around the horn…again! (all links go to the original Chinese article)

1) Tencent, A Manufacturing Re-evaluation

Summary: How does a company with social media and gaming roots try to transform manufacturing services? Good details on what makes Tencent different from other Internet giants trying to expand their domain, in particular the value of the WeChat ecosystem at the enterprise level.

Source: 机器之能/jiqizhineng (Synced) — a long-time source for ChinAI translations, often features longform articles about China’s tech industry

2) Tsinghua Law professor says “No” to facial recognition access control in neighborhoods districts

Summary: Reviews Professor Lao Dongyan’s protests against the use of facial recognition access control in entry/exit for neighborhoods/residential communities, which has become more common in a time of Covid. Also covers a seminar held on this topic held in Beijing on September 23. Recall that ChinAI #77 featured her strong case against the use of facial recognition on the Beijing subway.

Source: 南方都市报 (NDDaily) — Southern Metropolis Daily, newspaper published in Guangzhou -- well-known for its investigative journalism.

3) Right Brain AI

Summary: Longform article that examines Microsoft’s XiaoIce chatbot, as a new example of “Right-brain AI,” which enables machines to emotionally connect with human beings. Plays off the distinction with “left-brain AI” — which is more connected to calculation and rational decision-making.

Source:出色WSJ中文版 (Wall Street Journal, Chinese edition) — The WSJ’s Chinese-language site was launched in 2002. I wonder how often WSJ Chinese, or similar platforms like FT Chinese, translate their Chinese-language reporting into English.

4) “New Infrastructure” musters troops for the battlefield, BATH evolves toward intelligence

Summary: Assesses the positions of Baidu, Alibaba, Tencent, and Huawei in their attempt to construct the new infrastructure of the “Internet of Everything [万物互联].”

Source:钛禾产业观察 (Taihe Industry Observer). I covered this plaform in-depth in these two previous ChinAI issues, describing them as China’s mini DefenseOne. The first link has a translation of the 11th article in their “Transformer” series. This is number 15 in that series.

5) Yunxi assists enterprises in digital transformation with standardized digital “middle platform” products

Summary: To try and catch the wave of digital transformation, many companies have tried to implement a 中台 (zhongtai) strategy. This article is a brief look at Yunxi, a leading service provider of digital 中台 (zhongtai). I could not, for the life of me, figure out the best way to translate this term. Eventually, came across this Zhihu thread that suggested “middle platform.” An architecture concept first articulated by Alibaba, a middle platform is an additional connective layer that fits between the traditional 2-tier structure of business applications (back-end and front-end)

Source: 赛迪网 (CCIDNet). Interesting site that covers IT industry developments. Related to CCID consulting, which is also producing a lot of good content. I wrote a little bit about the output of consulting firms/think tanks like CCID, Qianzhan, etc. in ChinAI #72.

6) IDC Marketscape Report: “Assessment of China’s Conversational AI Vendors”

Summary: Released September 24, this report examines China’s dialogue-based AI market, which is predicted to reach 1.56 billion USD by 2024.

Source: International Data Corporation (IDC) is a Massachusetts registered and headquartered research company focused on the tech landscape. Interestingly, it was bought by China Oceanwide, a large Chinese conglomerate in 2017. IDC’s China division produces a lot of Chinese-language reports.

7) “AI Governance Online” Platform, launched in 2020 Zhongguancun Forum

Summary: This public service platform, a joint initiative by the Beijing Academy of AI and China-UK Research Centre for AI Ethics and Governance, was launched on September 19. Provides an evaluation for an AI project based on global AI governance principles as well as AI risk and governance case studies. You can play around with English-language version here; after, filling out the evaluation, it scores the project:

*It’s an impressive effort, and I applaud this important collaborative effort in the space of global AI governance. However, is this going to be a censorship-free space where the case studies section could include China’s use of facial recognition technology to target Uighurs and other ethnic minorities?

Source: 北京智源人工智能研究院 (Beijing Academy of Artificial Intelligence): BAAI was launched by the Ministry of Science and Technology and the Beijing Municipal Government in November 2018; it’s also supported by a lot of top Chinese universities and tech companies. For an interview with the head of BAAI’s AI ethics and safety research center, see ChinAI #52. For a summary of BAAI’s 2019 conference, see ChinAI #73.

8) An overseas PhD graduate at age 30, I decided to return to Shanghai Jiaotong University to Study AI

Summary: A profile of Jingwen Leng, who studied for a Computer Science PhD in the Department of at University of Texas at Austin, and has now returned to be a tenure-track Associate Professor at Shanghai Jiaotong University’s John Hopcroft Center. Relates his story to those of others in similar positions.

Source: AI科技评论(aitechtalk) — focuses on in-depth reports on developments in the AI industry and academia.

9) Is China’s Image on Twitter Getting Worse and Worse — A Tsinghua research team reveals the inside story

Summary: This article highlights this arxiv preprint that investigated China’s image among foreign publics based on a large-scale Twitter dataset. More interesting to me was the comment section, where some people noted the “guts/balls/courage” of Leiphone to cover such topics.

Source: 雷锋网 (Leiphone) — a long-time source for ChinAI translations, which I think of as China’s MIT Tech Review. See ChinAI #60 for more on this source.

10) China Standards 2035 vividly portrayed, this is the real key to Sino-U.S. tech competition

Summary: On glance looks like a good overview of China Standards 2035, albeit a bit hype-y and steeped in the great power competition narrative. Some good stats from Japan’s ICT tech committee on China’s growing influence in ICT standards.

Source: 智谷趋势(zgtrends) — new one to me, some harsh reviews on this Zhihu thread, describe themselves as cief intelligence adviser to decision-makers, and were one of Hurun China’s Top 50 most influential finance self-media.

***Okay same drill as last time: Reply to the email, or comment on the Substack post, with the number of the article you’re most intrigued by, and choose the feature translation for next week! If you’re supporting ChinAI financially as a subscriber, flag that in your reply, and I’ll give it a little added weight when tallying up all the votes.

ChinAI Links

Must-read: The Chipmakers — US Strengths and Priorities for the High-End Semiconductor Workforce

CSET analysis by Will Hunt and Remco Zwetsloot: “Technical leadership in the semiconductor industry has been a cornerstone of U.S. military and economic power for decades, but continued competitiveness is not guaranteed. This issue brief exploring the composition of the workforce bolstering U.S. leadership in the semiconductor industry concludes that immigration restrictions are directly at odds with U.S. efforts to secure its supply chains.”

Should-read: Scientists use big data to sway elections and predict riots — welcome to the 1960s

Harvard historian and writer of many New Yorker must-reads, Jill Lepore re-introduces us to the history of Simulmatics, which pioneered the use of computer simulations and pattern detection in American political campaigns. Draws from her book: If Then: How the Simulmatics Corporation Invented the Future (2020).

Should-read: Europe and AI: Leading, Lagging Behind, or Carving Its Own Way?

A really useful overview of Europe’s AI strategy (both at the European Commission- and country-level). Ending has some smart recommendations, and the paper frames Europe’s strategic position well:

• “Given the need to address the societal, ethical, and regulatory challenges posed by AI, the EU’s stated added value is in leveraging its robust regulatory and market power—the so-called ‘Brussels effect’—into a competitive edge under the banner of ‘trustworthy AI.’”

• “Yet normative principles and regulation alone are not enough for the EU to become a global AI leader. What is also required is a reevaluation of European competitiveness in this field in a way that leverages its comparative advantages… There is a clear rationale for a stronger EU-level role and for a more coherent European-wide approach to AI that complements member states’ own actions.”

Should-read: Xi's Science and Technology Speech Echoes and Updates Deng Xiaoping

Six experts on China’s S&T development untangle the past, present, and future in President Xi Jinping’s recent speech to a gathering of scientists and technology workers. DigiChina Project also translated the speech in full.

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 chinainewsletter@gmail.com or on Twitter at @jjding99