Andrew Greatorex At dawn, there was the neolithic agricultural revolution. Much later there was the industrial revolution. Slightly later still, the digital revolution came. With dusk, comes the machine learning revolution. Historically, dusk triggered caution in diurnal creatures like humans. And as the metaphor suggests, caution here should be respected.
Every major revolution marks a profound shift in society, often gutting historical economic and social norms to set up new precedents for a new era. In the face of such changes, public policy must adapt or face harsh economic truths.
The agricultural revolution, considered by many as the birth of modern humanity, was the transition from a hunter-gatherer lifestyle to one of agriculture, which would set the scene for societies to exist in built-up villages and towns, thereby providing the basis for division of labour, trading economies and early hierarchical and political structures. Whether this has been overall beneficial to humanity is debated — are we any happier today than we were when we were hunter-gatherers? (read Sapiens by Yuval Noah Harari). At any rate, the revolution changed humanity forever.
The industrial revolution powered industry through the transition of hand production to machinery, which meant that things could be produced exponentially more efficiently. Goods, news and people were all able to travel faster. Almost every aspect of daily life was changed. Average income and population began grow rapidly, as did standard of living. Importantly, the job market underwent a massive shift as machines took more and more jobs. Just as horses employed to carry cargo were gradually replaced by the car, and Turing’s Bombe machine replaced human decoders during the war, the industrial revolution brought machinery which put downward pressure on those in low skilled jobs. While innovation may be disruptive to jobs in the short term, in the long term there has often been little negative impact on the job market as a whole, since with innovation comes new opportunities for job creation. A similar story happened during the digital revolution, which proliferated the use of digital electronics in communication technology and computing.
The forthcoming machine learning revolution may be different. While ML technologies will provide unparalleled opportunities for business and society, it also poses a great threat to their job safety. Unemployment in the UK is at its lowest rate since 1975, but as (1) automation techniques continue to improve and (2) we continue to build ML models capable of doing medium to high skilled work in a fraction of the cost and time of humans, we will see this figure increase. And that’s just the difference. In previous revolutions, it was low skilled and repetitive blue collar jobs that were being replaced. And while the industrial revolution took place over a 50 year period and the job market had sufficient time to adjust, we are in the midst of an exponential transition in the creation and adoption of new technologies capable of dismantling the current employment model and displacing skilled professions in a short space of time.
The trend is inevitable and cannot be stopped (not that that’s necessarily a bad thing). Where we have fewer people who are employable, we must ask how we can deliver the benefits of an economy that effectively is beginning to run itself to those displaced workers.
At present, there are a few promising ideas in this area. The first is a concept known as “flexicurity” . Flexicurity is an integrated strategy to simultaneously enhance the flexibility and security of the labour market, hence its name. Healthcare, education and housing assistance would be provided to all, whether or not they are formally employed. But wait, wouldn’t this require some sort of magical money tree? In today’s age, it seems naive, but as more aspects of the economic model become automated, we may very well find a magical money tree to fund such projects. The main difficulty then becomes, will it be possible to fairly distribute the wealth created by machines or will they remain in the hands of the owners of production? Future public policy should reflect and prioritise this issue, for obvious reasons of growing wealth inequality. As well as flexicurity, we must mandate a guaranteed universal income. More focus should be placed on STEM education so that it can evolve faster to match the pace of innovation in order to develop a motivated workforce who are better equipped to hold down jobs.
So, how will people spend their time if traditional jobs are displaced? Will we have a generation of bohemian creatives who paint and sculpt and write music and poetry? Humans by nature require some sort of purpose, whether in the future that translates to creative or cultural pursuits or mass volunteering or something else only time will tell. Taking steps now in anticipation of the impending change is paramount, and will help people adapt to new economic realities.
Source: Towards Data Science