Also, AI Casting and the Netflix of China
|Jun 17 at 2:34 am||Public post|| 6|
Welcome to the ChinAI Newsletter!
*We’re close to 5000 subscribers — please share this with one person you think might like ChinAI if you find this content valuable. 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.
REBUTTAL TO THE NEW COLD WAR WARRIORS
In his last issue for Axios China, Bill Bishop — the “China-watcher’s China watcher” and creator of Sinocism (“the presidential daily brief for China hands) and one of the inspirations for ChinAI— headlined one of his final points as “A New Type of Cold War.” He commented, “Views are hardening among the elites in both countries. Trump clearly has bipartisan support in Congress for tougher policies towards China and the mood in D.C. is increasingly unforgiving to anyone who is not very hawkish on the PRC.” Technology and AI have become central to this “New Cold War” meme: My coauthors and I show in a Foreign Affairs article that before 2016 only a handful of articles mentioned the phrase “AI arms race”; today, an article on the subject gets published virtually every week, and Googling the term yields more than 50,000 hits. Some even warn of an AI Cold War.
These “New Cold War” Warriors, or Hankerers (the label I’ll give them) believe two things: 1) the current U.S.-China confrontation bears strong similarities to the previous Cold War between the U.S. and the Soviet Union, and 2) it is in the U.S. interest to escalate our China policy to a broader geopolitical context in which “Chinese influence” anywhere threatens U.S. interests everywhere. Each of these points are deeply flawed.
In this week’s ChinAI debate section, I challenge these tenets that drive the“New Cold War” narrative. A previous debate segment deconstructed Financial Times reporting that called into question U.S.-China collaborations in the AI space. In that piece, I mentioned that my inspiration from these segments came from experience in high school policy debate. When I debated, I was the “2N,” which meant that I gave the last rebuttal for the side of the negative, responsible for defending the status quo against the affirmative side, which would propose a plan of action. Perhaps partly influenced by years of fiercely defending the status quo, I argue in this week’s issue that those who hanker for a “New Cold War” have not sufficiently presented a vision that is preferable to the status quo. In other words, the so-called China Reckoning has gone too far.
On the first point, the current U.S.-China relationship is very dissimilar from the Cold War dynamic between the U.S. and the Soviet Union. The international system was bipolar then; it is definitively unipolar now. U.S. power is not only unipolar and durable but also underestimated especially in comparison to challengers such as China. U.S. military-technological superiority is not being eroded by China, and I have laid out a fairly comprehensive case for why U.S. national capabilities in AI far exceed Chinese ones in written testimony before the U.S.-China Economic and Security Review Commission.
During the Cold War, the American and Soviet economies were essentially two separate entities with limited connections in trade, finance, and activities by multinational corporations. The American and Chinese economies are intimately tied together through trade, finance, multinational production, and people flows amidst a more globalized world. Stop cherry-picking a few cases and generalizing about “decoupling,” and start looking at overview-type statistics that show how “coupled” the two countries are – say, the AmCham China 2019 China Business Climate Survey that shows how the technology sector reported higher earnings than other sectors, amidst a trend of increased revenue growth reported by American companies in China since 2015.
Yes, there are ideological differences between the two countries. I’ve pointed out how technology is linked to the horrible situations faced by Uyghurs in Xinjiang. But I’m not convinced that Chinese influence campaigns or the appeal of the “Beijing Consensus” has anywhere close to the potential to create Soviet-style satellite states. As Jessica Chen Weiss argues, a “fear that China represents a threat not just to specific U.S. interests but also to the very survival of democracy and the U.S.-led international order” simply “gets the challenge from Beijing wrong.”
What worries me most about ideologues and Cold War warriors is this collective forgetting of the dangers of the Cold War. Exhibit A: After the Cuban Missile Crisis in 1962, President Kennedy estimated that there was a one-in-three chance of nuclear war. One compilation lists seventeen close calls that could have triggered an accidental nuclear war. In other words, rerun history one hundred times, and humankind goes extinct in thirty-three of those simulations. I don’t know why journalists, policymakers, analysts, and academics are trying to promote this Cold War narrative, when we know that memes can be self-fulfilling, as Johnston’s research on China’s “New Assertiveness” meme demonstrates. Do we want to risk running back that Cold War history one more time?
If the situation doesn’t demand it (as I demonstrate in opposition to the Cold War agitators’ first point), there’s no need to adopt a contain at all costs approach (the second point of Cold War Warriors). In fact, overreaction is costly in many ways. It leads to McCarthy-like targeting of ethnic Chinese scientists, including U.S. citizens, searching for a cancer car as this Bloomberg account of what happened to Xifeng Wu, former director of the Center for Public Health and Translational Genomics at the University of Texas MD Anderson. In past ChinAI issues I’ve shared how this is an issue that hits home for me in particular, but these types of mishandled incidents only strengthen Chinese propaganda and recruitment efforts aimed at attracting high-skilled talent back to China.
The overreaction of Cold War hankerers are undermining U.S. interests in other ways. Graham Webster, in an insightful Twitter thread, points out that the U.S. approach to Huawei is not about protecting U.S. national security in a traditional sense — the president of the U.S.-China Business Council calls it akin to murder (of a specific company). Webster describes this approach as “a jump to containment without any kind of broader debate about its wisdom.”
Let this be my entry into the debate: maintaining the status quo is a defensible policy option to pursue U.S. national interests in technological and ideological competition with China’s party-state. Let’s flesh out what the status quo means. One version is maintaining the Obama doctrine on China that took hold in the years before the national disaster that is Trump’s administration, which has alienated allies and damaged the U.S. international reputation. Obama believed that the challenge posed by China’s rise required constant attention (hence the Pivot to Asia), and he backed this up with sensible policies that drew on multilateral coalitions: the opening to Burma, engagement with Vietnam and other Southeast Asian countries, TPP, attempts to integrate China with the international order, the recognition that a weakened and threatened China may pose just as much of a threat as a strong China.
Here’s a new “China reckoning.” We should be skeptical of any consensus that supports the Trump administration’s China policy — given that it is one crafted by former chief strategist Steve Bannon (who founded a platform for the alt-right and is known for directing white supremacist and Nazi beliefs into the mainstream), top trade negotiator Peter Navarro who doesn’t speak Chinese/has scant experience in China/and has a CV that “reflects someone with greater expertise in public utilities than the complex workings of the Asia-Pacific,” and Michael Pillsbury, who Trump falsely proclaims as “probably the leading authority on China” but somehow believes in a conspiracy theory that China’s leaders have had a secret 100-year plan to take over the world by 2049, one that has been thoroughly torn apart by the actual leading authority on China (in my opinion), Alastair Iain Johnston at Harvard.
When will the U.S. stop relying on old white men to shape its thinking toward China? And why has the left (which should know better given Trump’s regressive efforts toward every aspect of the left’s platform), the center (which should be holding), and the right (which should be wary of techno-industrial policy efforts) all seemingly acquiesced to the administration’s misguided approach to the world’s most important bilateral relationship?
Let me be clear — there are useful policy recommendations we can make and advocate for, especially investing in AI safety and preventing accidents from emergent and structural risks associated with advanced AI systems (I highlight three in my vision for U.S. AI policy here), but I think before we commit to any significant changes in the U.S. approach to China, we should seriously consider if the world would be better off if we just paused and re-assessed the situation. There’s always a bias toward action because we want to feel like we are doing something. You don’t win journalism awards for not publishing that piece that is going to spur overreactions and bad policymaking, but preventing a 5% loss is just as important as achieving a 5% gain.
In the debate over U.S. policy toward China, any affirmative side seeking to push productive policy must convince a negative side that offers a vigorous defense of the status quo. This type of deliberation will result in better policymaking. I’m happy to play the role of the status quo defender.
AI Chose Kris Wu to be a judge for “Rap of China.” What Can We Do About It?
This week’s feature translation is about iQiyi’s (the Netflix of China) approach to “AI casting.” It’s a fun, light read that gives a sense of some popular shows in recent years, such as the Rap of China. The author shows how a K-pop idol who had limited rap experience and a Cantonese pop singer who does not rap were chosen as judges by AI. Check out performances by two of my favorite Rap of China artists: Na Wu Ke Re who raps in English, Chinese, and Uyghur tili; and Wang Yitai, who just kills it on this song.
This Week's ChinAI Links
Back home in the greatest state of Iowa to play a bunch of ultimate frisbee and take it easy, and my public library had Broken Stars, an anthology of Chinese contemporary science fiction edited and translated by Ken Liu (who translated Liu Cixin’s The Three Body Problem) – it’s a must-read: the first piece is a fascinating one that recreates interactions between Alan Turing and a machine based on his childhood friend Christopher.
A rare week that features two must-read books: Matt Sheehan’s The Transpacific Experiment is available for pre-order. It chronicles the real human people who are making connections between China and the state of California.
Check out Danit Gal’s chapter that surveys on AI ethics efforts in China, Japan, and South Korea, forthcoming in the Oxford Handbook of Ethics of Artificial Intelligence.
Late to this one, but see Lindsay Gorman and Matt Schrader on the messiness of U.S. tech ties with China’s surveillance apparatus, in Foreign Policy.
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 email@example.com or on Twitter at @jjding99