“Artificial intelligence was not in popular demand before. I was left hanging out to dry before Apple Siri was released in 2011, or any breakthroughs were made in speech recognition.” Dr. Dong Yu, the former principal researcher of MSR (Microsoft Research) said. “Now, things have changed. AI, especially deep learning, is unbelievably hot.”
Dr. Yu just ended his 19-year’s career in Microsoft. Last Saturday, he was invited to give speech upon his latest research at the AI Next Conference in Seattle, WA. At the beginning of the speech, Dr. Yu said he joined Tencent, the maker of the mega-popular messaging app WeChat and QQ. His next move is to recruit Chinese AI talents and to establish Tencent AI lab in Seattle.
What people don’t know is the battle for smartest Chinese AI experts behind the scenes. It used to be the game between Silicon Valley’s big players like Facebook, Google and Microsoft. However, over the last few years, China’s tech conglomerates have unsurprisingly joined in the battle.
Chinese researchers and scientists are well-known for the greatest number of academic papers published on the subject. In 2015, 43% of the top academic papers related to AI were published with one or more Chinese researchers, regardless of where in the world the work was primarily conducted.
Meanwhile, Chinese AI experts in the U.S. played a big role in deep learning frameworks. Dr. Yu was the main contributor of CNTK (Microsoft cognitive toolkit), an open-source deep learning toolkit from Microsoft, for example. Caffe, a deep learning framework made with expression, speed, and modularity in mind, is created by Yangqing Jia, a Chinese Ph.D. in computer science at UC Berkeley. He is also the co-author of Google’s Tensorflow. Another open-source deep learning framework MXNet is created by a group of Chinese computer scientists.
However, Chinese AI talents are still in shortage, said Dr. Yu. “AI was being largely ignored and underfunded before the deep learning boom. There are so few people who have thrown themselves into this field with a long-term commitment. Many of my colleagues left AI research and moved to other fields like search and ads.”
He estimated the number of top-tier AI researchers – like Geoffery Hinton from Google, Yan Lecun from Facebook, is only about 30 to 40 at all (he was not sure about the number. It is definitely no more than 100).
Besides, the rapid rise of AI industry drives an extremely huge demand for talents. Dr. Luming Wang, the head of Uber deep learning, showed me an example. Uber’s self-driving program has to process a big collection of 10-million-mile data run by Uber drivers every day, which needs a big number of data scientists and researchers to handle the issue. The demand for AI scientists will boom if artificial intelligence is integrated with vertical industries like agriculture, retail and energy, revolutionizing almost every aspect of our economy and human life.
“Once a trustworthy and successful model is accepted and adopted at different vertical fields, there might be a hundred times or even a thousand times of the current demand for AI talents because each field needs people to do deep integration and deep application. It might happen in 2 or 3 years at all.” Dr. Wang said.
Tech Giants are Hoarding AI Talents
Well-established tech companies are still the first-tier choices for Chinese AI academics, especially Google, Microsoft and Facebook, said Dr. Yu.
Google’s deep learning program has been generating incredible amounts of buzz over the last few years since the acquisition of Deepmind, the maker of AlphaGo and other breakthroughs. Google is investing heavily in AI, recruiting an AI research team of 250 people, led by noted AI expert like Geoffrey Hinton, said Dr. Yu. Besides, Google announced last November that they will extend funding for AI research at the Montreal Institute for Learning Algorithms, one of the best academic AI institutes in the world, led by well-known AI expert Yoshua Benjio.
Facebook’s AI lab Fair is somehow less influential, consisting of about 30 AI researchers, mostly young academics. Remember Yangqing Jia, the creator of Caffe? He was the research scientist of Google Brain since 2013. Last year, he joined Facebook and became the research and engineering lead.
“Fair is attractive to young academics because it is considered to be one of the best firm institutes for academic research.” Dr. Yu said.
MSR used to be the best tech company in attracting Chinese AI talents. Until now, Microsoft still has many Chinese principal researchers like Li Deng, Xuedong Huang and Jinyu Li. However, MSR is experiencing a reshuffle of executives and the brain drain.
Speaking of industrial applications and tools, Amazon and Uber are performing pretty well in talents recruitment. Dr. Wang told me about 40 percent of the faculties are Chinese in Uber deep learning platform team.
Chinese Conglomerates are Coming for Chinese AI Talents.
The west coast used to be the battlefield for Silicon Valley’s big players. However, China’s tech giants like BAT(Baidu-Alibaba-Tencent), Huawei and Xiaomi are joining in the battle as well.
In 2014, Andrew Ng left Google and became the chief scientist at Baidu, a Chinese search-engine company often likened to Google, working at the company’s research lab in California’s Silicon Valley (He recently resigned from Baidu). Earlier this year, Qi Lu, former Microsoft executive joined Baidu as the new COO. Alibaba, the world’s biggest online retailer, has also launched an AI lab in Seattle, recruiting over 60 people, mostly Chinese AI researchers, from companies like Microsoft, Google, Amazon, etc. And don’t forget Dr. Yu.
China’s tech company is becoming an appealing option for Chinese AI researchers.
Big package (money, stock shares and other bonus) is a big factor why Chinese smartest minds are lured away from companies like Microsoft and Google. Companies like BAT are always lavishing attention, resources and money on AI talents. Sometimes, the cost to acquire a top AI talent equals to the cost of getting a NFL quarterback, according to rumors.
Big data is another big advantage. Due to the comparatively lax data privacy law in China, Chinese companies are rich of data, which is useful for deep learning.
China is leading the development of virtualization and digitalization. Chinese companies and are eager to invest once AI experts present them with the opportunities to grow their business or make bigger savings. For example, the digital payment industry is booming in China that creates opportunities of integrating banking and artificial intelligence in order to handling massive data created from user end.
Author: Tony Peng, Synced Tech Journalist