Recent waves of technological innovation will have a deep impact on the labour market due to their rapidly expanding ability to perform human tasks and the declining cost of implementing them. Some of the most disruptive advances occur in the field of artificial intelligence (AI), where machines learn to perform tasks in a similar way to humans.
Hence, AI-powered machines are already making headway in doings tasks once considered out of their reach, and solely in the realm of robots in science fiction.
But where machines win, some human lose. The question, of course, is which one of us is on the losing side.
According to Carl Benedikt Frey and Michael A Osborne, researchers from Oxford University in England, while a vast amount of jobs are susceptible to being automated by AI, three critical areas remain out of their reach.
The first one is tasks that relate to perception and the manipulation of objects in unstructured environments such as housekeeping, where numerous objects are left lying about in no particular pattern.
The second one is tasks that involve creativity such as solving novel problems. And the third is tasks that rely on social skills such as caring for other people.
Therefore, jobs that require higher proficiency in at least one of these critical areas – take psychiatry as an example – are unlikely to be replaced by machines in the foreseeable future. On the contrary, data entry jobs are very likely to be replaced since the occupation doesn’t require proficiency in any of those three areas.
There are now 8.3 million Thais working in high-risk occupations, or jobs where the estimated likelihood of them being replaced by AI is over 70%. A little over half of those who risk losing their jobs to machines hold a middle school diploma or lower educational achievement. This may create the impression that a higher level of education and some technical skills can guarantee a secure job.
However, a number of tech companies and startups around the world are proving this statement to be wrong. SMACC, a German startup, is set to disrupt accounting departments around the world by developing a machine learning software to take care of various financial tasks like invoice data extraction and processing, optimised payments, and even liquidity analysis.
In addition, several legal tasks can now be accomplished by AI. Robo-Lawyer, for example, is an AI-powered automated system that provides clients with legal advice. Other innovations are automated contract due diligence, legal analytics, and a class of systems which gathers and screens for data relevant to the case called eDiscovery.
As demonstrated, all levels of jobs will soon be affected by automation one way or another.
When change is inevitable, the only option is to adapt. In the book Only Humans Need Apply , authors Thomas H Davenport and Julia Kirby propose five ways that humans can survive the first onslaught of AI.
The first strategy to accomplish this is to “step up” to fill the gap left by robots, which are highly efficient in specific tasks but unable to relate different tasks to form a coherent picture of the overall objectives and goals. This strategy typically involves moving up to high-level executive roles.
The second route is to “step into” AI and the automation development process. Such jobs obviously include engineers and software developers but corroboration from other domain experts is also instrumental in bringing new technology to a working application.
While most people are not experts, they can still employ the third strategy by “stepping in” and embracing AI automation to enable one to perform tasks once deemed infeasible or too resource consuming.
In addition, people who have a particular interest in some niche area can also “step narrowly” into fields considered too small to justify the kind of sizeable investment required to develop an AI automation system.
Lastly, economic values are created when people interact with other people; hence, there is still vast room for people to “step aside” from robots and unleash their uniquely human skills, namely, empathy and emotional intelligence.
That all seems fine in theory, but how exactly are we going to follow these five steps? Let’s take lawyers, an occupation already disrupted by AI as demonstrated.
First of all, a lawyer can “step up” to become a case project manager who oversees the whole litigation process. Secondly, a lawyer can “step into” the legal automation development process by, for example, encoding her or his knowledge of law into the form which computer can utilise. Other lawyers can “step in” by using AI to help craft a litigation strategy using analysis on a large number of related cases.
Laws permeate most aspects of human society, hence it is inevitable that there must be many narrow but obscure fields still far from AI’s reach.
Some lawyers will be able to profit by “stepping narrowly” into these fields, such as an esoteric but profitable equine law. Finally, trial lawyers are obvious examples of those who “step aside” and emphasise human skills. Lawyers who can develop a close relationship with clients would be another example of lawyers who “step aside”.
That is not to say deploying those strategies would be easy, nor would the transition be painless. While some may be well positioned to take advantage from advances in AI, such as programmers or tech savvy lawyers. Others may not be so fortunate; and this is where government interventions will be crucial.
While in some cases it might make sense for the government to take direct action in reskilling its workforce, such as providing training courses, the modern economy is too complicated for such an initiative to be feasible on a large scale. A better alternative would be to accelerate reskilling by promoting the competitive training market. Such a measure includes, but is not limited to, grants and subsidies for private training providers, a training cost subsidy for workers, and a comprehensive information system for training courses.
These measures have been implemented abroad and the regime would be wise to imitate them. Any failure to do so would diminish the competitiveness of our nation, thereby reducing growth and prosperity, while at the same time threatening social cohesion by dividing people into those who can take advantage of AI and automation technologies, and those who can’t.
Nuthasid Rukkiatwong and Kotchakorn Khwamchareon are researchers at the Thailand Development Research Institute (TDRI). Policy analyses from the TDRI appear in the Bangkok Post on alternate Wednesdays.
First Publish: Bangkok Post on Wednesday, August 08, 2018