Blog: Why AI will most badly affect white-collar, not blue-collar, jobs
Moravec’s paradox explains why white-collars, not blue-collars, will be hit first by AI
Most people, when thinking about the consequences of AI, believe that new technologies will most badly affect manual labor workers and blue-collar jobs, while high-skilled jobs like accountants will remain unscathed. However, so-called Moravec’s paradox challenges this assumption and, contrary to conventional wisdom, states that blue-collar jobs will be less affected by AI and robotics than white-collar.
This principle was first advocated by Hans Moravec, Rodney Brooks, Marvin Minsky in the 1980s. As Moravec put it,
it is comparatively easy to make computers exhibit adult level performance on intelligence tests or playing checkers, and difficult or impossible to give them the skills of a one-year-old when it comes to perception and mobility.
There are biological reasons behind this phenomena, and Moravec attributed this paradox to the theory of evolution. According to this explanation, skills that have evolved over millions of years, even before Homo Sapiens appeared, (motor and social skills, for instance) are much harder to be mimicked by new technologies, whilst those evolved over the past several thousands of years (mathematical and scientific thinking) are relatively easy to be replaced by AI algorithms.
Otherwise stated, the difficulty of reverse-engineering any human skill is proportional to the amount of time that skill has been developing. What is easy for humans (moving, interacting with others) is hard for AI; and what is hard for us (complex analysis of information and subsequent predictions and decisions based on it) is easy for AI algorithms.
As one expert has said, “In essence, AI is great at thinking, but robots are bad at moving their fingers”.
Therefore, it is far easier for algorithms to spot patterns and make predictions based upon data (for instance, financial analysis) than for robots do ordinary housework. This is the reason why computers like IBM’s Deep Blue already outperform humans and why we still do not have robots doing am housework for us.
Algorithms are easy to distribute, adapt, and improve.
Robotics, however, is much more difficult. It requires a delicate interplay of mechanical engineering, perception AI, and fine-motor manipulation. These are all solvable problems, but not at nearly the speed at which pure software is being built to handle white-collar cognitive tasks. Once that robot is built, it must also be tested, sold, shipped, installed, and maintained on-site. Adjustments to the robot’s underlying algorithms can sometimes be made remotely, but any mechanical hiccups require hands-on work with the machine. All these frictions will slow down the pace of robotic automation. — Kai-Fu Lee
In the past, during Industrial Revolutions, physical automation hurts blue-collar workers first, but in the forthcoming decades, it will be the white-collar jobs which will most negatively be affected by AI.
In fact, McKinsey’s study finds that the demand for “office support” jobs (financial workers, IT workers, administrative assistants) will drop by 20% in the US by 2030, whilst demand for “unpredictable physical work”(machinery installation and repair workers, agricultural field workers) is expected to grow by 6%.
This doesn’t mean that blue-collar jobs will not be automated. They will be replaced too, but the paces of elimination will be slower than those among white-collar due to peculiarities of AI and robotics.
In any case, we should begin preparing for the unsettling future now in order to be able to withstand the sweeping changes of the 21st century.