David Wei's employer, a Beijing-based artificial intelligence start-up, doubled his previous salary and offered share options to entice him to come and work on the company's AI-powered software for the legal industry.
That was eight months ago. The 27-year-old programmer says he already regrets not asking for more money – he currently earns 400,000 yuan ($86,977) a year not including options – and is now actively looking for a better-paying gig, the South China Morning Post reported.
"My friends advised me to stay put and complete at least a full year so that it won't look bad on my resume," said Wei, who has a postgraduate degree in applied mathematics from Harbin Institute of Technology, one of China's best-known universities, along with Tsinghua and Peking University, when it comes to science and engineering.
But with more than 10 recruiters calling him with job offers since the start of the year, the advice from his friends is looking increasingly outdated.
Wei is one of the estimated 300,000 people in the world who have the required skills in advanced computer science that underpin developments from driverless cars to home robots.
The global demand for AI experts like this is in the millions, according to a report by Tencent Research, a unit of China's Tencent Holdings.
The AI talent shortage is expected to be one of the issues that comes up for discussion at the annual closed-door Central Economic Work Conference this month, where members of the ruling Communist Party and government department heads gather to discuss how to implement the vision for a technologically advanced nation set forth earlier this year by Chinese President Xi Jinping.
The conference sets the course for key economic policies over the next year and delegates will likely discuss measures needed to achieve the targets set out in July when the State Council issued a roadmap to develop an AI industry worth over 1 trillion yuan by 2030.
Dubbed the fourth industrial revolution, the rise of AI – powerful supercomputers crunching data to identify patterns and teach themselves tasks – has fuelled concerns over its potential to put people out of work.
One estimate by management consultancy McKinsey forecasts that 800 million jobs worldwide are at risk of being lost to automation by 2030.
In China, technology companies like Alibaba Group Holding, which owns the South China Morning Post, are experimenting with unmanned convenience stores and warehouses, and using AI to determine credit scores for Chinese consumers.
JD.com, the online retailer, just announced an investment into driverless vehicles and is using drones to deliver parcels. Insurers and banks are using AI technology to power their ATMs and improve the customer risk-assessment process.
All of these artificial intelligence initiatives, ironically, require an abundances of human brainpower, which is why people like Wei are in such demand.
As of March, there were more than 1.9 million technology professionals working in the AI sector globally, according to LinkedIn statistics. While the US accounts for the lion's share with 850,000, China ranked seventh with just over 50,000, trailing India, the UK and Canada and on par with Australia and France.
China's demand for AI professionals may surge to 5 million in a few years, People's Daily reported, citing Zhou Ming, an education vice-director at the Ministry of Industry and Information Technology.
So far, China has been importing most of its AI talent from overseas because its home-grown AI talent remains scarce, according to LinkedIn's Global AI Talent report published in July.
Roughly nine out of 10 AI positions advertised on the mainland go unfilled, unless companies venture out of China to hire, according to the professional networking platform.
The number of China-based AI job postings on LinkedIn surged more than eight times over the past three years, from about 50,000 in 2014 to 440,000 in 2016, said Wang Di, vice-president of LinkedIn China.
About 44 per cent of the overseas AI talent working in China come from the US, followed by the UK and France as the second and third source countries, according to LinkedIn.
Researchers in machine learning, smart chips and algorithms are the most sought-after, according to Wang. Demand will continue to grow, with a survey by global risk advisory company Willis Towers Watson showing that Chinese tech firms project an average 20 per cent expansion in their AI teams next year.
"Many Chinese tech companies now visit top universities and headhunting firms on a regular basis, keeping an eye out and learning about the market. Their overseas recruitment plans would even have no limit when it comes to headcount or budget," said Willis Towers Watson managing director Edward Hsu.
In terms of recruitment strategies, Hsu said Chinese companies first try poaching AI talent from leading peers as well as hiring fresh graduates from home and abroad.
Longer term, they are setting up overseas labs to attract foreign researchers and expatriate Chinese unwilling to return to China, while some are looking at outright acquisitions of competitors as a means to acquire the talent, he added.
Indeed, a Silicon Valley presence is now seen as standard procedure for any mainland Chinese company serious about working in AI.
Car hailing company Didi Chuxing, China's answer to Uber, opened a US-based research lab in March after successfully poaching seasoned cybersecurity experts, including a former Uber researcher known for being able to hack cars with a laptop computer from miles away.
Tencent opened its own AI lab in Seattle, Washington state, in May while Alibaba pledged in October to invest more than US$15b on AI research and development over the next three years with a plan to open seven labs worldwide aimed at beating similar research facilities built by IBM, Microsoft and Intel.
Shortage of talent may become the biggest barrier for China to achieve its AI dreams, said James Lewis, senior fellow at the Centre for Strategic and International Studies, a US-based think tank.
"Its AI road map is good, but China may need more time to build the workforce," he said, adding that nurturing and growing a talent pool in technology can be a slow process.
Even so, China's economic strength and willingness to spend money gives it an advantage over the US or Europe, Lewis said.
It is coming down to competition between US and Chinese companies for the small pool of talent, and the recent Chinese "brain drain" is being reversed now that good AI scientists are being offered a "fair amount of pay" to work in China, he said.
With over 700 AI-related companies in China, about US$2.6b of capital has flowed into the industry over the past four and a half years up to June, roughly one seventh of the amount going into the US but three times that of the UK, according to a joint report by Wuzhen Institute and NetEase.
For Wei, the AI programmer in Beijing, the talent shortage is good news.
"When I land at the next start-up, a 30 per cent pay rise will be the least I'll get, although I could offer some discount if it's already a tech giant," he said, when asked about his salary expectations.
However, coming from a remote village in northeast Shandong province, Wei admits to being shy about discussing his salary, even with close friends, in a country where he earns almost 40 times the average wage.