A.I. and the Job Sector: Demon 2.0?

Post By RelatedRelated Post

By Philip Apfel 

In the 18th century, when the industrial revolution meant that newly created machines were bound to render obsolete or at least fundamentally redefine many jobs, developments in the world of machinery were often met with borderline apocalyptic cris de coeur by the anxious citizen. Thomas Carlyle, a well-known satirist, philosopher and historian at the time, spoke of the “demon of mechanism” annihilating the modest workman’s source of livelihood; destructive and cruel. With the benefit of hindsight, we now know that he was mistaken. Though it was no doubt a bumpy ride, technological innovations ended up creating more jobs than they destroyed, heralding an unprecedented era of economic growth and prosperity.

Today, ground-breaking developments in artificial intelligence throw up similar questions. Only this time, history might not repeat itself, as the challenge that AI poses to human employment is both multi-sectoral and more difficult to adjust to than it was in Carlyle’s days.

According to a 2013 study published by Frey and Osborne of Oxford University, about 35% of current UK jobs are at high risk of being taken over by computers in the coming 20 years. Considering that the service sector makes up more than three quarters of the UK economy, this number is striking. Across the pond, a very recent report by ‘Forrester’ states that by 2021, robots will have eliminated 6% of all jobs in the US, with employees in the customer service and transportation industries affected first. Virtual assistants that can be used in the customer service industry, like Apple’s ‘Siri’ and Microsoft’s ‘Cortana’, are becoming ever more sophisticated, and a fleet of driverless cars, hitherto only to be found in sci-fi movies like “I, Robot”, was recently let loose onto the streets of Pittsburgh by transportation network company ‘Uber’.

High levels of investment into AI companies, to the tune of almost $9 billion in 2015, and sophisticated new and revamped AI techniques, such as the idea of “deep learning”, have helped propel the industry to a level where present machines are already indifferent to the colour of our collars.

In medicine, some machines developed by San-Francisco-based start-up ‘Enlitic’ already outperform expert radiologists in the identification of tumours from x-rays. In accounting, a German tech start-up called ‘Smacc’ has managed to develop cloud-based accounting software which can already provide for most accounting needs that small and medium-sized enterprises might have. In the legal sector, doc-review tasks can already be carried out by machines using predictive coding technology. This technology allows machines to read, analyse and distract pertinent information from thousands of documents in a much shorter time than the human eye. In Brown v BCA Trading 2016 EWHC 1464 (Ch), London-based law firm Berwin Leighton Paisner recently won the first contested application to use predictive coding as part of a document review exercise. Within the next few years, use of such technology could well become common practice. Finally, and perhaps most impressively, AI is now also making inroads into journalism, an area that at first blush would seem to be one of the least susceptible to a take-over. Appearances can be deceiving. Artificial Intelligence from the American tech company ‘Automated Insights’ researched for and wrote 1.5 billion newspaper articles last year. For all you know, a machine might have written this very Beaver article.

Leaving aside the seemingly limitless efficiency and productivity opportunities that AI-developments in all these sectors represent, it is clear that these developments herald considerable challenges for the labour market of the coming years. Challenges which dwarf those of our assiduous forefathers. Whereas the move from agriculture to manufacturing was a process that took many years, giving people time to adjust, sophisticated AI software can be developed quickly, and brought into use even faster. A bank can start using algorithms executing equity trades from one day to the next, rendering the employee who made them previously redundant in just as short a time. Furthermore, disruptions are no longer limited to a particular sector of the economy. Instead, they span most sectors. As a result, escaping to a sector in which one can still add value as a human being will be more difficult than it was for our predecessors.

If current progress in Artificial Intelligence continues at such a rapid pace, we may soon have to fundamentally rethink the way we organise ourselves as a society. Perhaps the notion of a ‘job’, where you ‘work’ every day will become obsolete. That may sound quite attractive to some. But then there is the uncomfortably dystopian possibility of all future gains of a robot-driven economy accumulating amongst the few, without proper distribution of wealth. That sounds much less attractive.

Alternatively, one can always go into philosophy. Or acting.They probably won’t take those away just yet.  

Features Editor