DEC 3 marked the opening of the annual World Internet Conference in Wuzhen. The global techfest was attended by the CEOs of some of the world’s largest companies, including Apple’s Tim Cook and Google’s Sundar Pichai, alongside Alibaba Group Holding’s Jack Ma, Tencent Holdings’ Pony Ma and Baidu’s Robin Li, as well as the top names in artificial intelligence (AI), robotics, digital payments and cybersecurity. The opening session on the first day was aptly titled “Artificial Intelligence: Smarter World, Better Life”.
That a small Chinese town 90 minutes’ drive from Shanghai would host the world’s biggest annual internet event, or that CEOs of global tech giants and the world’s top robotics experts would flock there, should not surprise anyone. China is where the action is across an array of powerful internet-enabled technologies such as AI, robotics, fintech and e-commerce as well as virtual and augmented reality.
The real battle is for AI dominance. Just how important is AI? Russian President Vladimir Putin put it succinctly in a speech to Moscow students in early September: “Whoever takes the lead in AI will have the key to global domination.” Though Putin is probably under the illusion that Russia itself might be in the race, there are really only two major contenders for the world’s first AI superpower: China and the US.
Winner takes all
Why does AI matter so much? Because like much of the rest of the tech sector, AI is likely to emerge as a winner-takes-all technology. Few of us can recall who the second-biggest e-commerce player in the US is behind Amazon.com, or the distant-second player in the search-engine space behind Google, or the second-largest social network behind Facebook.
A US government report last year noted that “AI holds the potential to be a major driver of economic growth and social progress”. China too sees AI as the new growth driver as it enters the next phase of development. As the country’s demographics change for the worse with an ageing labour force and declining productivity, it needs to leverage technology such as AI to power future growth. In July, China unveiled its “New Generation Artificial Intelligence Development Plan” with the stated goal of being on a par with the world’s finest by 2020, making AI the primary driver for Chinese industry by 2025 and becoming a world leader by 2030. Consulting firm PwC estimates that AI’s impact on productivity could boost China’s GDP by up to 26%, adding US$7 trillion to its total gross domestic product, by 2030.
As it is aggressively deployed over the next 12 years, consultancy McKinsey & Co estimates AI’s contribution to China’s overall GDP growth could be 1% to 2% annually.
To be sure, China is clearly several years behind the US in AI. Yet, it is leaving no stone unturned in its efforts to catch up, with the Chinese tech giants and Beijing itself pouring tens of billions of dollars into AI. “China understands that it needs to invest more heavily across the AI ecosystem if it wants to be a global tech leader,” says Catherine Wood, CIO of ARK Investment Management, a New York-based investment group focusing on disruptive technology. The slower China’s march in AI, the bigger the lead US tech giants — Amazon, Google’s parent Alphabet, Facebook, Apple and others — can build, making it harder for China and its tech companies to eventually catch up.
What China has going for it is its long-term focus and its ability to get its state apparatus, universities and Chinese tech giants to collaborate for a greater national goal. “China sees the AI race as a marathon, not a sprint,” says Wood. “Trust me, it won’t be long before China overtakes the US in the development of advanced AI,” Eric Schmidt, chairman of Alphabet told a tech conference in the US last month. Schmidt believes at the current rate, it would be five years before China starts to catch up.
All about data
The key to AI is data. “There is no data like more data,” AI pioneer Prof Fred Jelinek wrote when he first laid out the promise of the technology. With a population of 1.37 billion, China has nearly 900 million smartphones in use, or about four times that of India or the US. A typical person in China carries his or her smartphone around all day and it collects a lot of personal data from all the apps installed, notes David Dai, China technology analyst at Sanford C Bernstein in Hong Kong. More-over, companies such as Alibaba, through its e-commerce platform, Baidu, through search, and Tencent, through messaging, music, video and gaming, are collecting enormous amounts of raw data, indeed even more than Google, Facebook and Amazon collect in the US. This is in part because the Chinese are messaging more, buying more stuff online and making more payments digitally through AliPay or WeChat Pay every day — whether they buy chestnuts from roadside stalls or a dress from a department store — than their American counterparts.
Since data is retained both within the app and the system, vast amounts of it can be extracted from smartphone users. The handset tracks calls; the browser tracks what people are searching for; the e-commerce engine knows what they are are buying, how much they are spending and where; the sensors track the number of steps the user takes; and the GPS chip tracks where they go and when. In offices and homes, the home speakers are listening to every word, even when they are switched off. “Data-privacy rules in China are a lot more liberal than they are in Europe and the US, which allows Chinese tech companies to harness big data and AI a lot more powerfully than their counterparts around the world,” says Ajay Kapur, Asia-Pacific strategist for Bank of America Merrill Lynch in Hong Kong. “Chinese citizens are not against having their data collected, as government monitoring is a fact of life in China,” notes Bernstein’s Dai. That gives Alibaba, Tencent, Baidu and others the sort of competitive advantage that their developed-world counterparts can only dream of.
China’s sheer size gives it a major advantage as well. It has more than 200 million students in its school system. That means 200 million sets of work and exam results to analyse, notes Dai. In healthcare, China has many more patients, which means that it has enough samples even for some of the rarest diseases, he notes. In e-commerce, Alibaba’s US$560 billion gross annual sales tower above Amazon’s US$136 billion revenue. Countries and companies that generate more data have an advantage over those that generate less as they seek to boost their AI capabilities.
Within AI, China has picked image recognition, machine vision, speech recognition and natural language processing as key niches, notes Dai. In image recognition, the country is already a world-class player. In NLP and speech recognition, China enjoys a natural advantage because of the language difference, the Bernstein analyst says. Apple’s Siri, Amazon’s Alexa and Microsoft’s Cortana recognise our voices and search the internet through clever use of AI software. Amazon’s and Google’s home digital assistants are powered by AI. Alibaba is experimenting with using AI to mark Chinese language essays at Zheijiang University. Within five seconds, Alibaba’s software identified errors in the essays, including redundant words and wrong selection of words.
Wooing top global talent
Chinese tech giants such as Alibaba, Tencent, Baidu and JD.com are paying top dollar to woo the world’s best AI and robotics engineers and coders from Europe, North America and India, enticing them with not just hefty salary packages and perks but with the option to buy some of the world’s best-performing tech stocks. Software engineers from Bangalore and San Francisco know that they could be within reach of the proverbial jackpot when they sign up with a Chinese tech company.
These days, a 30-year-old AI engineer with three to five years of industry experience can command up to US$500,000 in annual salary, in addition to bonuses and perks. China needs all the AI talent it can get. According to some estimates, there are only 50,000 experienced AI professionals in the country, compared with more than 850,000 in the US.
A White House report published just a week before the US presidential election last year noted that China now publishes more research on deep learning than the US, while AI-related patent submissions from Chinese researchers have increased 200% in recent years. Last year, China increased its output of AI-related papers by almost 20% over the previous year even as output from American and European researchers dropped. Yet, while China leads the world in the quantity of AI research, it still lags behind the US and Europe in the quality of fundamental research.
Beijing and the Chinese tech giants are aware that they need to improve, so besides luring the best AI experts to China, they are buying AI start-ups in the US and opening new AI research centres in the US as well as teaming up with US companies to do research or, in some cases, help fund the research. Among the most aggressive in AI is Tencent, which had been seen as a laggard because Baidu and Alibaba have been pouring billions into the technology for years. Tencent’s big push into AI has included poaching some of the most talented researchers in the US and setting up a new AI lab in Seattle, which will be run by former Microsoft principal researcher Yu Dong. Tencent also recently poached Zhang Tong, a machine learning guru who previously headed Baidu’s big data research lab, Huang Zhiheng, a speech recognition expert who was Facebook’s top research scientist, and Liu Wei, who previously headed Chinese ride-sharing giant Didi Chuxing’s machine learning lab.
Alibaba also runs an AI lab in Seattle in addition to its labs in Hangzhou. JD.com has an R&D centre in Silicon Valley and Didi Chuxing recently opened a research lab focused on AI and autonomous driving just down the road from Google’s headquarters in Mountain View, California. Baidu, which this year lured top Microsoft AI expert Qi Lu, has extensive AI research operations in China and the US.
One indication of China’s intense focus on AI has been the introduction of AI-enabled home digital assistants or speakers by tech giants including Alibaba, Baidu, Tencent and Xiaomi. One company that has differentiated itself from the pack is iFlyTek, which uses AI to power image and voice recognition systems that help doctors with their diagnoses, aid teachers in grading tests and let drivers control their cars with their voices. Another firm that has used AI effectively is Hangzhou Hikvision Digital Technology Co, the world leader in surveillance camera systems, which has come under attack in the US for its close links to the Chinese military. Hikvision sold surveillance cameras to US municipalities as well as the federal government and is ironically now under surveillance itself as to whether its devices collect data that is being passed on back to the Chinese government.
As China closes the gap with the US in AI, Washington is likely to watch the large Chinese tech companies more closely and prevent the transfer of sensitive technology. Three years ago, US regulators upended a bid by Tsinghua Holdings, a Chinese tech group, for US memory chipmaker Micron Technology. In 2015, the Obama administration blocked Intel, Nvidia and Advanced Micro Devices from selling high-end supercomputer chips to the Chinese. And in September, President Donald Trump blocked the sale of Lattice Semiconductor, which has programmable software technology to build AI chips, to a Chinese firm on national security concerns. Indeed, the biggest gap between the US and China right now is that the former has all of the cutting-edge chip technology to strengthen its AI and robotics ecosystem while China has none.
China will continue to pour billions into semiconductor research and is eventually likely to close the gap. Until then, it will remain focused on its inherent data advantage to advance its audacious AI game plan.
Assif Shameen is a technology writer based in North America