ChinAI #220: NEC Silicon Valley - the Lab that Dragged Chinese Companies into the AI Era
Greetings from a world where…
this Kings-Warriors series is magical
…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: NEC Lab, the one who dragged Chinese companies into the AI era
Context: If your starting point for thinking about China’s AI landscape is a geopolitical chessboard in which technological assets map onto airtight national containers, that’s okay — personally, that just feels like such a flat and narrow way to comprehend this topic. What if we start with a different perspective: China’s AI ecosystem as made up of people, with dreams and aspirations, who flow across borders even as they run up against structural barriers.
This week’s feature translation (link to original Leiphone article) traces China’s AI development back decades to a crucial wellspring: the Silicon Valley lab of NEC (a Japanese multinational information technology and electronics corporation). Through this history of Chinese AI researchers at NEC, we meet Yu Kai (who later led Baidu’s institute of deep learning), Wei Xu (now chief scientist of General AI at Horizon Robotics), and many other people who played key roles in cultivating China’s AI ecosystem. When you read it, I hope it makes you seek out more multidimensional and broad ways of thinking about this topic.
Key Passages: NEC Silicon Valley was essential to Baidu’s “All in AI” pivot. Let’s go back to 2012, when Baidu CEO Robin Li met up with Yu Kai, who directed the NEC lab and had long been a devout believer in deep learning.
After Google fully withdrew from the Chinese mainland market, Baidu was flushed with success. Robin Li, who is sensitive to technology, kept up with the dynamics of top technology teams at home and abroad, and soon realized that AI would bring new opportunities to Baidu, and began to look for talents everywhere. When the two met, Yu Kai preached on deep learning, and Robin Li talked about his interest in speech recognition and image recognition. There was the surging excitement of two people who regretted not having met earlier.
It was also on this day that Yu Kai and Robin Li made a historic mutual choice for Baidu and China’s AI landscape: Robin Li threw out an olive branch and invited Yu Kai to return to China…Previously, Baidu’s research accumulation in AI was almost zero. It was only after Yu Kai joined that Baidu took the lead in injecting deep learning genes. In just a few months, Yu Kai sold Robin Li on the potential of deep learning…Initially, Baidu supported the first batch of companies in China that did data labeling.
On a secret bidding war for Geoffrey Hinton’s services:
However, what really made Robin Li determined to "take a heavy position" on deep learning was the failure of an auction that Yu Kai participated in on behalf of Baidu in December 2012. The neural network AlexNet developed by Geoffrey Hinton's team won the world's top image recognition competition ImageNet in one fell swoop, causing a sensation worldwide. Yu Kai was deeply affected. As a former champion, he knew better than anyone about the disruptive nature of the AI technology paradigm represented by AlexNet. So, he immediately sent an email to Hinton, expressing his eagerness to cooperate.
This email inspired Hinton to launch a secret auction at the NIPS conference that year, attracting four world-renowned leaders in the field of AI: Google, Microsoft, Baidu and DeepMind. Baidu bid tens of millions of dollars, but eventually lost to Google. But through Yu Kai's eyes, Robin Li witnessed the madness of the powerful opponent of the United States in pursuing deep learning, and felt the crisis.
In January of the following year, Robin Li prepared to establish the world's first AI laboratory named after "deep learning" – the Baidu Institute of Deep Learning (IDL)…
Yu Kai was the actual operator of Baidu IDL in the early days. Under his influence, Baidu attracted a large number of outstanding AI scientists, such as Wei Xu, the founder of the PaddlePaddle platform, Chang Huang, the co-founder of Horizon, and Yuanqing Lin, the dean of Baidu Research Institute. Coincidentally, they have all visited the same place in the past: Silicon Valley’s NEC Lab.
In the 1980s, when the Japanese economy was thriving, NEC established two research labs in the United States, one at Princeton and another in Silicon Valley. How do we get from NEC Silicon Valley to China’s AI ecosystem today? We start with Yihong Gong, who introduced the first group of Chinese AI researchers to NEC, including Wei Xu and Shenghuo Zhu who interned at NEC while completing a doctorate in machine learning at the University of Rochester.
Gong was one of the first batch of China’s government-sponsored students who went overseas to Japan. He went to the University of Tokyo to study in 1983…After graduation, he briefly taught at Nanyang Technological University in Singapore and Carnegie Mellon University (CMU) in the United States, but only for a few years. Ultimately, by a random opportunity, he landed at the media analysis department of NEC Lab in Silicon Valley.
In mid-August 2005, Zhu went to Salvador, Brazil, to attend ACM SIGIR, the top conference in the field of information retrieval. It was at this meeting that Zhu Shenghuo met Yu Kai for the first time. At that time, Yu Kai was twenty-nine years old and had just graduated a year ago from the University of Munich with a Ph.D. degree…At this conference, apart from doing academic exchanges, Zhu also planned to recruit some talents to enter NEC. He had read Yu Kai's work before, and he was very impressed. He happened to meet him this time, so he went up to chat with Yu Kai. When talking about future opportunities, Zhu asked Yu Kai: “Are you interested in going to the United States?” Yu Kai was planning to change jobs, and he got along very well with Zhu Shenghuo, so he responded that he was very willing.
Zhu treats people warmly, and Yu Kai is also very talkative. The two quickly became acquainted in Brazil. Amidst all of this, Shenghuo Zhu was robbed and got hurt from the struggle with the thieves. In admiration, Yu Kai praised him for his courage, but he smiled and waved his hands and said, "It’s nothing. When that person robbed me, I just kept telling him to take whatever he needed, but he didn’t understand me.” Yu Kai thought, “This guy is so silly, so fun.”
However, during that time, Yu Kai was also considering returning to China to advance his career, and even got an offer from Microsoft Research Asia…the most famous AI research institute in the world. At that time, NEC Labs America was also very strong. There were many top aces at the master level, such as Vladimir Vapnik, the inventor of SVM (Support Vector Machine), and Yann LeCun, the pioneer of deep learning.
Yu Kai hesitated several times, and finally, with the advice of his friends, he decided to go to Silicon Valley first. Yu Kai, who entered NEC in 2006, was determined that he would eventually return to China. No one expected that this idea, six years later, would directly promote the transformation of the entire Chinese AI industry.
First half of FULL TRANSLATION: The past happenings of NEC Lab in Silicon Valley: the one who dragged Chinese companies into the AI era
*We’ll finish up the second half next week!
ChinAI Links (Four to Forward)
Should-read: DigiChina Translation of China’s New Draft Rules for Generative AI
For DigiChina, Seton Huang, Helen Toner, Zac Haluza, and Rogier Creemers (with editing by Graham Webster) have produced an English-language translation for “Measures for the Management of Generative Artificial Intelligence Services (Draft for Comment).”
Should-read: Decoding China’s Ambitious Generative AI Regulations
Sihao Huang and Justin Curl, both studying at Tsinghua University as Schwarzman Scholars, have published a good analysis of the above measures. Their take, in the “Freedom to Tinker” blog, hosted by Princeton’s Center for Information Technology Policy:
We were often surprised by the CAC’s apparent willingness to impose considerable costs of compliance on service providers who use generative AI models. If fully enforced, provisions mandating strict pre-training data controls, security assessments, model optimization, and de-biasing would be incredibly expensive. This expense may slow adoption and progress in deployed, public-facing AI systems. Yet just as complying with China’s cybersecurity laws required Chinese tech firms to build extensive content moderation systems, these new rules may incentivize companies to invest more in building robust, interpretable, and aligned generative AI systems, albeit ones that espouse the country’s values.
Should-read: AI is already taking video game illustrators’ jobs in China
For Rest of World, Viola Zhou reports on how recent breakthroughs in AI image generation have displaced jobs in video game illustration:
“I wish I could just shoot down these programs,” the artist told Rest of World, after getting off work late one night. She said fear of impending layoffs had made her colleagues more competitive; many stayed at work late, working longer hours to try to produce more. “[AI] made us more productive but also more exhausted,” she said.
Should-attend: Book Launch: Deter, Disrupt or Deceive
On May 2, if you’re in DC, the American University’s Center for Security, Innovation, and New Technoloyg is hosting a book launch for Deter, Disrupt or Deceive, which studies cyberspace activities more as an intelligence contest (less like a war). The panel discussion features four of the book’s contributing authors: Richard Harknett, Nina Kollars, Jon Lindsay, Erica Lonnergan, and Joshua Rovner.
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 an Assistant Professor of Political Science at George Washington University.
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