|May 13||Public post|| 1|
Welcome to the ChinAI Newsletter!
These are Jeff Ding's (sometimes) weekly translations of writings on AI policy and strategy from Chinese thinkers. I'll also include general links to all things at the intersection of China and AI. Please share the subscription link if you think this stuff is cool. Here's an archive of all past issues. *Subscribers are welcome to share excerpts from these translations as long as my original translation is cited.
I'm a grad student at the University of Oxford where I'm based at the Center for the Governance of AI, Future of Humanity Institute.
If I I could rewind time and rewrite last week’s issue, I would have focused on the larger systemic issues/context surrounding why reporting on these issues is so hard rather than targeting the reporting itself. As many people reminded me, the debate format can sometimes optimize for "“winning” over more productive deliberations. The goal should have been to build each other up to do better rather than take each other apart over mistakes. In parts of my rebuttal, my sarcastic (and let’s be honest — douchey) tone was unfair to both the journalists and commentators, and detracted from the substance. I apologize to all injured parties and promise to learn from my mistakes. That doesn’t mean I won’t express my strong opinions on issues (see my musings on Chinese Americans as strategic assets in the last section of this issue), and I stand by all the substantive arguments in my rebuttal. It just means I need to recognize the immense privilege of having a platform like ChinAI, as well as the privilege of being a male in this space (when I go in on an issue, I don’t need to worry about being labeled “emotional” or “ranty”), and do better on all these fronts.
One last clarification on the reporting details from last week, and then we’re moving on. I respect the FT reporters and staff for defending their reporting, as Madhumita did in her responses to some of my points, which I linked at the bottom of last week’s post. Thanks to FT for correcting the article to include links to the coauthored papers. Madhumita informed me that one of the papers (the one on how facial features, dress, and voice collectively affect the human sense of beauty) originally shared with me was not one of the three FT was reporting on. Instead it was this paper on machine reading comprehension.
I do want to address one point of substance: why am I harping on the technical content of these papers? I think it links to bigger issue with AI+politics/tech+politics research — or what I call the “technology abstraction problem.” An exercise to illustrate this: take a sample of AI+politics articles/papers that claim AI has revolutionized X, and replace AI with “high-level statistics.” The best policy research on AI should use the word artificial intelligence in an abstract sense as few times as possible. AI has become too prone to hype and it’s too broad of a concept to be analytically coherent or useful, encompassing anything from subfields of fuzzy mathematics to research on decentralized drone warfare. Analysts should rigorously force themselves to specify what claims they are making about “AI” in terms of the domain and technological layer they are talking about.
4 million Chinese Americans as Strategic Resources for the U.S. and China
This week’s two translations are not on spectacular research analysis or incredibly well-reported investigative pieces but rather they are “thermometer” pieces that give a good sense of the overall temperature of Chinese public/influencer opinion on a particular subject. The subject in question is the position of Chinese Americans as a strategic asset for the U.S. and China in the competition over scientific and technological development.
The first piece, by Dong Jielin, an associate researcher at China Institute for Science and Technology Policy (CISTP) of Tsinghua University, who received her PhD in Physics from Carnegie Mellon University. She states that Chinese Americans “Are very important strategic resources for China and the United States,” and if the “ ‘friendly-to-China’ Americans and the ‘friendly-to-America’ Chinese are purged clean, and the two sides are divided clearly into two camps, then the bridge is broken and there will be no one to repair it.” She reviews the history of Chinese Americans in the United States and highlights the cases of American physicist Xiao Xiaxing and U.S. Meteorological Administration expert Chen Xiafen as instances of discrimination toward Chinese Americans in wrongful charges of espionage. It’s a short piece so I’d recommend reading it in full, but one tidbit I found particularly interesting. Dong references that many people in China accuse Chinese immigrants in the U.S. of causing frictions in the U.S.-China relationship due to their misconduct (e.g. illegal tech transfer) and Dong calls for Chinese Americans to “strengthen their legal awareness and compliance.”
The second piece shows how some Chinese thinkers and media view U.S. targeting of Chinese American scientists and engineers as a strategic opportunity. Usually in ChinAI issues, I feature the “cream of the crop” Chinese reporting/analysis, so "these “thermometer”-type pieces give a truer sense of how there are still a lot of flaws in the media environment. Titled “The United States has extensively restricted ethnic Chinese talents, the backbone of technology has returned to China, and the edge in AI has been reversed,” this piece gets basic facts wrong and spreads some scary rumors.
One passage states, “Washington believes that all Chinese-American talents are inherently suspect as ‘spies.’ The White House even uses its power to go to major high-tech companies and other companies involved in key technologies, and ordered that them to remove all these Chinese talents within a certain time period to ensure that US technology will not flow out.” Unless I’ve been living under a rock (to be fair, have not been checking Twitter all that much lately), this is a false rumor, presumably, because the author wants to push the narrative that the US is no longer friendly to Chinese-American scientists and engineers and encourage them to move back to China.
Reflections on Chinese Americans as “Strategic Resources” in the US-China Tech Competition
Chinese Americans are not strategic assets. We are people. Translating these two pieces reminded me of this video of the announcement of the 2008 Nobel Prize in Chemistry which was awarded to Roger Tsien along with two others — I would strongly recommend watching it alongside my breakdown of what happens. At 13:40, a journalist from Chinese news agency Xinhua asks, "Are you Chinese? Can you speak Chinese?" Roger's response is pure grace. Born in New York to Chinese immigrants, he responds in Mandarin, "I can speak a bit." Xinhua follows-up by asking what his achievement means to Chinese scientists. “Well I can't say I’m really a Chinese scientists. I grew up entirely in America, but if this should make Chinese people feel good and proud and it will inspire a lot of young people to do science in China, that's a good thing." It's an eloquent reply that takes claim of his own identity but also one that acknowledges why the reporter is asking the question: he recognizes that no Chinese citizen of the PRC had won a Nobel Prize in science, and that aspiring Chinese scientists may draw inspiration from someone who looks like him winning.
Still, what I can't shake about this clip is the drawn-out "ummm" after he's being asked if he's a "Chinese scientist.” I wonder what was flashing through his mind in that "um." Was he thinking about the astronomical odds that his mom faced in immigrating to the U.S. when the Chinese Exclusion Act was in force? Was he recalling how a NJ developer wouldn't sell to his family because they were Chinese? Was he thinking about how for so many Chinese Americans, no matter how good your English is, how high you climb (even getting a Nobel prize), there will be some people who only see you as Chinese? In a world where science is more politicized, increasingly framed in the context of international competition, where you hear whispers about people of Chinese ethnicity not having the U.S. national interest in mind, Roger's story reminds us that Chinese Americans have agency to claim their own identities, that we who have Chinese faces and family but U.S. passports and roots contain multitudes like everyone else, but inhabit multiple identities more often than everyone else. That it matters to read Roger's story to see a great Chinese American scientist who did work for the world. Roger passed away in 2016. He was a fine pianist, a gifted amateur photographer, and loved by family, friends, and colleagues. He was a person, not an asset.
I think this links back to why I was so mad about those FT pieces. I wonder if FT saw the humanity in the Chinese American researchers involved in the collaborations, if they would have tried to get the views of people who would have defended the collaborations or at least added more context. One of the researchers named in the FT pieces was Thomas Huang, who, like me, was born in Shanghai. In 1949, his family moved to Taiwan where Huang studied electronics before going on to get a Doctor of Science at Massachusetts Institute of Technology (MIT). I don’t pretend to have any insight into Professor Huang’s political views, but let’s just say that most people with families who moved to Taiwan at the time when Mao proclaimed the founding of the People’s Republic of China are usually not schills for the Chinese Communist Party. Huang has taught at the University of Illinois at Urbana-Champaign, where he was named as a Swanlund Chair, the highest endowed title, for his contributions to one of those midwestern public research universities (shoutout to the University of Iowa, my alma mater) that are the engines of America’s competitiveness and economic strength.
Again, as I emphasized in last week’s newsletter, Huang’s collaboration with Cloudwalk on re-ID research is concerning. What I’m trying to get at here is that as we can critique personal actions, government policies, and the CCP, let’s not lose sight of the humanity of Chinese people. Recently, I was at one of those meetings where a bunch of powerful “influencers” get together and talk about “influencer” things. I gave a brief on China’s AI development alongside a Chinese American friend. After the briefing, I was surprised by how many people came up to me and confused me with my friend (we look nothing alike). It reminded me of a passage from a Vulture profile of Korean American actor John Cho, which has been stuck in my head ever since July 31, 2016, when I shared it in my previous (now-temporarily-defunct) newsletter project, the New Chimericans, a collaborative effort with my good friend Laura Wang that looked at Asian American issues. John makes this comment on Asian American representation in media:
“I’ve seen many instances where we’re seen as a little less than human, or maybe a little more than human — like ultrahuman, rather than subhuman. What is wrong with film representation? Some of it is mechanical, surprisingly. I’ve thought about why Asian stars — from Asia, I mean — look so much better in their Asian films than they do in their American films, and now I can answer that to some extent. There’s an eye, and it’s not a malicious eye, which is a way that the people working the camera and behind the scenes view us. And then they process it and they put it on film. And it’s not quite human. Whereas Asian films, they are considered fully human. Fully heroic, fully comic, fully lovely, fully sad, whatever it is. And it’s this combination of lighting, makeup, and costume.
If you don’t think of a person as fully human, you sort of stop short and go, That’s good enough. Do you remember Doug Liman’s film Go? I remember Taye Diggs in that movie, and he was charcoal black. I was surprised to see him in “How Stella Got Her Groove Back” — I realized that Go was not an accurate representation of his skin tone whatsoever. And I’ve met him. He was carelessly lit. Why is that? Why is one carelessly lit? The white people were carefully lit.”
How do we tell stories, write analysis, and understand the role of Chinese Americans in U.S.-China strategic competition that sees Chinese Americans as "carefully lit” human beings. Recently, there were reports that the U.S. State Department’s policy planning staff is composing a memo in the style of George Kennan’s “X Article” that argues the coming conflict with China is “the first time that we will have a great power competitor that is not Caucasian.” As Professor Ward points out in analysis for the Washington Post, the argument Trump’s State Department is making is not about ideology or civilization. “It is about race. China — unlike Russia — is not predominantly white, and thus must be dealt with differently.”
I think Chinese Americans can play a valuable role in dissecting these dangerous modes of thought. We represent an America that is not “Caucasian.” We have a “strategic in-betweenness.” Many of us are high-skilled, Western-educated Chinese natives who can move back and forth between the countries. But we can also leverage this “strategic in-betweenness” to question the humanity behind arguments like the ones coming out of the U.S. State Department about race-based great power politics.
This Week's ChinAI Links
Chinese phrase of the Week: 泛泛之辈 (fan4fan4zhi1bei4) - a mediocre person.
Human Rights Watch report on a system that “surveils and collects data on everyone in Xinjiang. The system is tracking the movement of people by monitoring the “trajectory” and location data of their phones, ID cards, and vehicles; it is also monitoring the use of electricity and gas stations of everybody in the region.” Also, read this account of women fleeing Xinjiang and telling their stories.
Show some love for the American Mandarin Society newsletter - they are a great resource for those looking to keep up their Mandarin abilities.
Check out and subscribe to DigiChina’s must-read monthly digest: an exclusive from April’s edotopm — translated excerpts from Zhang Shu, a researcher at the China Information Technology Security Evaluation Center (CNITSEC), analyzing U.S. “containment” strategies in technology and arguing China should prepare for a “long-term competition by promoting openness.
ChinaLawTranslate is an invaluable resource for unofficial translations on Chinese legal documents. See this new translation of Measures for Determination of Violations of Laws and Regulations in APPs' Collection and Use of Personal Information (Draft)
Thank you for reading and engaging.
Shout out to everyone who is commenting on the translations - idea is to build up a community of people interested in this stuff. You can contact me at firstname.lastname@example.org or on Twitter at @jjding99