人工智能時代的工作變革

喬·盧克·巴恩斯

The future of labor in a digital economy has become a heated topic alongside the emergence of artificial intelligence （AI）. Will the proliferation of AI and robotics herald human redundancy as a working species？

This prospect has inspired pessimism in many Western countries that is not shared by most Chinese. Indeed， according to a survey compiled by the Dentsu Aegis Network1， just 18 percent of British and German citizens feel that new digital technologies will create job opportunities over the next five to ten years. In China， a country with a labor force of around 800 million， 65 percent of people believe that AI will create even more work.

Government backing

A major factor is that few governments are embracing the digital age with as much gusto as China. In July 2017， Chinas State Council set a national goal of becoming the worlds primary AI innovation center， aiming to foster an AI industry that produces in excess of one trillion yuan （US$147.7 billion） by 2030.

Such enthusiasm has seeped right down to the municipal level， with local governments especially keen to support startups in the sector.

“The business environment in China and especially Shanghai is very attractive，” says Stéphane Truong， founder of Actionable Data， an AI consultancy service company. “I have seen a lot of initiative from several city districts such as collaboration with incubators2 to propose ancillary services， organizing competition for financial subsidies and providing a flexible fiscal policy.”

But Beijing has perhaps created the most fertile environment for tech startups. The capitals Zhongguancun area is known as Chinas “Silicon Valley” due to concentration of tech startups based there. Its proximity to Chinas two premier academic institutions， Peking and Tsinghua universities， makes it a happy hunting ground for new talent.

Disappearing jobs

One startup is Oriental iFly， which aims to use AI to create an automatic grading system for essays that provides instant feedback to teachers on students work and saves time spent marking.

I asked one of the companys product designers， Kailin Xie， whether this innovation might put teachers out of work.

“Teachers arent hired to grade，” she asserts. “As long as there are students， teachers will be necessary. Grading is just an extraneous part of the job. Our product enables a teacher to save dozens of hours a week on marking essays.”

Is a school likely to pay those teachers the same for less work？ Or will it instead use those extra hours to give them more classes， which would reduce personnel requirements？

Such questions could be some of the defining issues of the digital age. Should companies use AI to increase productivity and profits， or do they also have a duty to improve the day-to-day routine of their employees？

Business optimism

Much of the tech community has adopted the belief that these problems will simply sort themselves out. This is certainly the attitude of Stuart Leitch， founder of Lollipop.ai， a Seattle and Shanghai-based software company that uses AI to improve customer engagement3 with online products. “Firms have a very bad habit of hiring for unnecessary positions. The employees arent bad， but their duties usually involve repetitive， brainless and low-value work.”

“We want to release people from those positions and reduce the cost of that kind of work so those people can do more meaningful things. At the end of the day， we expect to create jobs across industries rather than put people out of work，” explains Leitch.

But what of the manufacturing jobs that have served as the backbone of Chinas economic growth？ Many are likely to go， admits Denny Xu， vice president of the Shanghai Haihe IT Company， which produces intelligent speech robots.

“AI will change future employment trends and patterns，” he explains. “Now， the labor force is too costly， so lower-level labor will largely be replaced by AI-related technology. But humans wont be completely unnecessary—human-machine coupling will become a future trend for enterprises and businesses.”

Skills gap

To a large extent， the challenge is retraining people. Thousands of brand new jobs are being created. In fact， a growing complaint from business leaders and recruiters is a lack of talent with the necessary skills to fill emerging jobs.

Stuart Leitch notes， “Its especially difficult to find talent on the development front. Since the skills most in-demand for our business are hard data4 science and machine-learning skills， were finding that we need PhD-level candidates， which are few and far between. It may become necessary to just hire go-getters who can learn quickly but dont necessarily have experience.”

Stéphane Truong at Actionable Data has similar issues. “My company would ideally employ candidates with at least a masters degree in computer science. We offer a very competitive package， but the battle for talent is rough because employees are more attracted to mature companies.”