What Jobs Are Affected by the Development of AI?

Job security is high on the list of concerns for people looking for a job. A study by the Institute of Policy Studies found that more than half of all Singaporeans worry about job security and employment prospects.


Moreover, we are living in a world where machine learning and artificial intelligence (AI) is developing so rapidly, it’s revolutionising the way we work.


In this article, let’s look at what jobs will be affected by these developments in AI.


AI isn’t the first major revolution to the way we work. Back in the 18th century, people thought the Industrial Revolution would put most people out of work.


The Industrial Revolution transformed major industries like agriculture, textile, transportation and metallurgy, greatly decreasing the need for farmers, manual labourers and craftsmen. Yet today, we look back at the inventions of machines to do the work of hand tools and the factory system as indispensable.


Will the revolution of AI have the same effect on the world?


McKinsey Global Institute calculates that intelligent agents and robots could eliminate as much as 30% of the world’s human labour needs, displacing the jobs of up to 800 million people by 2030. Eventually, 99% of all work as we know it can be done by machines.


But before you start panicking, know that these developments in machine learning and robotics will not happen all at once.


Replacing humans with AI takes a lot more than just technical feasibility to happen.


The four other factors required for automation to occur include:


  • The cost to develop and deploy the hardware and software for automation,

  • The related supply-demand dynamics and labour-cost for the work activity,

  • The performance of automation beyond just labour-cost (i.e.: quality, output, errors),

  • And regulatory and social-acceptance considerations.


Most importantly, McKinsey’s research findings simply show that work activities people are paid to perform can be replaced by robots, but a majority of occupations will not cease to exist entirely.


The net effect of AI isn’t not no jobs for people to do but a transformation in the way we work.


Highly Affected Jobs

Sectors like manufacturing, hospitality and retailing are most susceptible to automation because a large majority of the work involves physical activity or operating machinery in a predictable environment.


But this is based on technical considerations alone.


Book-keeping, accounting and auditing jobs that are highly automatable cannot entirely be replaced with AI because human judgement and emotional intelligence is required to use this data processed by machines to make decisions.


However, such administrative jobs are more likely to be automated compared to jobs in the food industry because the skills and training required for administrative work is scarcer than basic cooking, thus it is more cost-effective to automate book-keeping than food preparation.


Work activities in retail jobs like maintaining records of sales and customer information and stocking and packing merchandise for sale or for shipping can be easily automated with AI but the social aspects of retailing like advising and interacting with customers requires emotional intelligence that only humans can provide. Thus, only a small percentage of the work of a salesperson can be replaced with AI.


Moderately Affected Jobs


Up to 43% of all work activities in finance and insurance jobs can also be automated with AI since half of the work done in this sector is dependent on data collected and processed.


Insurance sales agents gather customer information, security and financial sales agents require data to prepare sales and other contracts. Even stock traders and investment bankers who earn an annual income of more than $200,000 currently spend a large amount of their time doing those things.


These data collection processes will eventually become automated in order to free up time for people with finance and business occupations to spend more time advising clients and making important decisions.


Other types of jobs that will be moderately affected by AI are those that require large amounts of physical activity or the use of machinery in unpredictable environments. For example, work in sectors like farming, construction and emergency services involve a lot of physical work that can be replaced by machines.


But since the work environment of these jobs are unpredictable and frequently changing, it is not easy for all of the work activities in these occupations to be replaced by AI. However, with continual advances in technology and machine learning, this is bound to change.


Least Affected Jobs


Finally, work activities that involve managing and developing people, decision making, planning and creative work will be least affected by AI.


Computers perform better than humans in collecting and processing data but humans are still required to determine goals and interpret results, so for sectors where human interaction is part of the work like healthcare and education, it is unlikely for these jobs to be replaced by AI.


While AI has provided us with the possibility of predicting diseases and accurately diagnosing patients, the most important work done by nurses and doctors involves communicating and interacting with people.


And finally, teaching jobs in education cannot be fully automated because the essence of teaching requires human expertise and complex interaction with other people.


So while administrative work activities like collecting student particulars and monitoring academic progress can be done by machines and technology has enabled classes and learning to occur online, teaching is still very much the work of humans.


A World of AI & Humans


Driverless trains are already a reality today and autonomous cars will soon be deployed for public use. The developments of AI technology these days mean that the demand for driver jobs might be eliminated entirely very soon.


Fully autonomous cars are theoretically safer and more efficient than human-driven cars, however, discussions regarding the regulations for autonomous cars are still ongoing and getting everyone in the industry to use the same reliable machine learning algorithm remains a big hurdle.


Some day in the future, AI will eventually become so developed that machines can maintain themselves and upgrade their systems autonomously. Perhaps McKinsey’s predictions might come true and 99% of all the work we do today will be done by AI in the future.


But even so, there will always be jobs for us to do and ways we can occupy ourselves. After all, people play video games for a living today—a ‘job’ that only became a job after the proliferation of social media.


Life as we know it might change, but occupations, jobs and the industries that make up our society will continue to exist. What’s more important is that we embrace the developments of AI and invest in the education and the skills required in a world of AI and humans.

Leave a Reply