How Will Automation and AI Impact the Labor Market?

The world of work is continuously changing, and ways of automating menial tasks has had an impact on almost every facet of society. Sometimes this has been to improve quality, by removing the imperfections resulting from manual human handling. Other times it is to improve productivity, resulting in the loss of jobs as we are displaced by machines.

In order to understand how AI and automation will impact the world of employment and productivity, the McKinsey Global Institute (MGI) undertook a two-year study to determine the potential impact on the global economy.

One starting point was to determine the capability requirements for certain activities within clearly defined occupations. In total, over 800 occupations were assessed, encompassing more than 2000 different activities. Each activity was analyzed to understand what it entailed. This ranged from cognitive capability, such as logical reasoning and problem solving, to physical capabilities, such as dexterity and motor skills.

The report indicates that around 60% of occupations have around 30% of activities that are automatable. Less than 5% were considered to include activities that are 100% automatable. The reason these numbers are so high is because the automation capability is more intelligent than ever before, in large part due to improvements in AI.

Unlike previous periods of technological improvement, the next wave of automation implementation will have an impact across the board. This is in contrast to, as one example, the 20th century where the impact of automation was mostly felt within the agriculture sector.

Tasks that are most likely to be passed onto automation solutions include the collection of data, processing of data, and activities with predictable physical traits. These would include physical activities and operating machinery in predictable environments. This transformation touches both low-skilled and high-skilled employment roles.

The study hopes that, like previous technology-enabled shifts, increases in automation will not necessarily demand long-term mass unemployment. As has been seen before, new opportunities usually arise that were not foreseen. For example, any automation solutions will require development, programming and maintenance. Thus, for automation and AI to deliver any efficiencies or improvements, it must be undertaken hand-in-hand with human labor.

Another key point to note is that automation will not happen overnight. AI has exceptional power, but it is still only being trialed in many business sectors. In industrial settings, AI has to be coupled with machinery and robots in order to be useful. Such hardware takes a lot longer to develop and certify in comparison to pure software products. As such, the potential for progress is naturally slower.

Further factors to consider are economic and regulatory acceptance. The automation solutions must prove to be economically viable, whilst society, especially governments, need to ensure that appropriate regulation is in place in some cases. This is of utmost importance where lives could be put at risk, such as for healthcare or transportation.

Automation and the potential provided by AI will have the broadest impact compared to any previous technology-led improvement. The study also indicates that half of today’s work activities could be automated by 2055 but, equally, this could occur 20 years sooner or later. Certainly, many roles will change, and not only unskilled roles. As mentioned earlier, loss of labor in some segments will be offset by new jobs created in other markets. Finally, we have plenty of time to plan and prepare, as long as we remain aware and open to the changes that are happening around us.

Arago Redaktion 2. May 2018