Will Al eliminate half the job?


  • 27 February 2020

Employment demand outweighs substitution

In my opinion, it may not reach such a large proportion. Moreover, the development of artificial intelligence will certainly have a "replacement effect" on existing jobs, but it will also produce a "demand effect" and "innovation effect" that increase employment. And the intensity of the latter two will be greater than the former.

As artificial intelligence technology improves labor productivity, the wages of on-the-job workers will be increased, product prices will be reduced, and corporate profits will be increased, thereby stimulating the consumer demand of on-the-job workers and the investment needs of enterprises, and thus expanding the scale of employment .

The "innovation effect" of artificial intelligence is even more prominent. Creepy engineers, cloud computing experts, online consumer anchors, short video producers, and other common occupations today are unpredictable ten years ago. In the future, artificial intelligence will continue to restructure production and consumption methods, and in the next ten years will spawn a large number of new products, new industries, new formats and new employment.

Will Al eliminate half the job?

AI shortcoming

Although artificial intelligence technology is good at many tasks that humans despise, human's effortless ability has become its shortcoming. For example, human beings can quickly summarize regular intuition from small data, understand social skills, recognize everyday concepts and scenes, and so on.

Some of human's seemingly simple capabilities are actually supported by the human brain's 6 billion neurons and 500 trillion connections. A person's brain capacity is equivalent to 2 to the 50th power byte, which is equivalent to the current capacity of the entire Internet. I dare not pretend that computers can never simulate the full capabilities of human beings, but at least for ten years, the boundaries of AI capabilities are limited.

Structural employment changes not a nuisance

Of course, thinking about structural employment changes triggered by artificial intelligence is not a nuisance.

First, there may be a shrinking of middle-skilled jobs. There is no doubt that high-skilled people who are good at using artificial intelligence to increase professional value will gain an advantage in employment. For lower-skilled positions, especially in the field of life services, which requires flexibility, situational judgment, and communication with people, artificial intelligence is not capable in the short term.

However, as artificial intelligence becomes more and more adaptable to repetitive tasks such as computing, retrieval, and classification, as well as the flatness of the hierarchy due to digitization of enterprises, the vulnerability of middle-skilled jobs that had undertaken routine programming has increased. Highly repetitive and low-innovation professions, such as bank tellers, face transformation pressure.

Will Al eliminate half the job?

Demand for IT professionals is rising sharply

A classmate of the head office of a national joint-stock bank told the author that in 2019 they plan to recruit 1,200 IT graduates, and there is very little demand for economics and finance graduates with a purely liberal arts background. The middle-skilled people are the cornerstone of the middle-income class, and the social effects of the reduction of middle-skilled positions deserve attention.

Second, the comparative advantage of men and women in employment may change. Automation technology reduces the physical requirements for employment, the platform economy improves the flexibility of employment, and women's participation in the labor market is gradually lifted.

At the same time, while artificial intelligence technology has replaced a large number of male physical employment, the comparative advantage of women in medical care, wholesale and retail work is gradually emerging. The population of women in the Tokyo metropolitan area has grown significantly faster than men over the past decade.

Bring opportunities

Furthermore, regional employment gaps may be more pronounced. Artificial intelligence technology will promote the concentration of employment in megacities, while employment in small and medium cities, mainly manufacturing industries, will decline. McKinsey's analysis of more than 3,000 counties and 315 cities in the United States found that the labor force of 54 backward cities and about 2,000 counties where a quarter of the U.S. population is growing older, smaller, and higher in unemployment Automation technologies have amplified these trends. The author's research on the employment structure of various provinces in our country also found that the proportion of manufacturing employment in the central and western regions reached a peak below a low level, and there were hidden concerns about premature deindustrialization.

Finally, the globalization of artificial intelligence brings opportunities. The time difference for applying the latest technology in different countries has been greatly reduced. Computers entered Vietnam 11 years after the birth of the United States, and today Github's latest open-source code sharing speed is only momentary.

Will Al eliminate half the job?

Humans are more focused on service

With the acceleration of the globalization of artificial intelligence, especially the maturity of "remote intelligence" (RI) technologies such as machine translation, telepresence and augmented reality technologies, and innovative collaborative software platforms, the globalization of service industry supply may be realized in the next ten years. When Chinese workers can wear augmented reality AR helmets to enter working scenes thousands of miles away, remotely operate unmanned machinery through the touch function of sensors, and communicate with employers through machine translation headsets, "Made in China" is upgraded to "Chinese Services" It is not out of reach.

Artificial intelligence technology is racing against institutional change. Artificial intelligence will ultimately benefit the public in the long run, but short-term costs should not be overshadowed by a vision of long-term prosperity. The disappearance of old jobs may be faster than the industrialization of new technologies, and the transformation of workers in specific fields between different jobs is not easy.

Public policies

At present, the speed of technological change has been ahead of the adjustment of public policies. Artificial intelligence is becoming an important issue in policy discussions in various countries. The lifelong learning and employment training system that adapts to the globalization of artificial intelligence, is conducive to human capital investment rather than taxation that favors technology capital investment. Policies, etc, are all options for future public policy.