It is now a decade since the publication of 'Race Against The Machine' by Erik Brynjolfsson and Andrew McAfee. This is seen as the seminal work that highlighted the possibility of the increased use of robotics and automation leading to a significant displacement of labour in the economy. Some of the wilder forecasts suggested that up to a half of all jobs could be lost to automation and AI. We are now a decade further down the road, which gives us an opportunity to take stock of how accurate some of those forecasts have been.
How far down the road could we expect to be in ten years? The main factor underpinning this view is the operation of 'Moore's Law'. This establishes that the speed and power of computers doubles roughly every two years. The way this works is that, over a decade, if Moore's Law still holds, the speed and power of computing would increase not by five times, but by a factor of five. Such is the power of the exponential. Whilst there have been significant increases in the speed and power of computers in thet past ten years, to suggest that this has happened by a factor of five is a bit of a stretch.
The key to increased automation and use of robotics is the further development of Artificial Intelligence (AI). It seems that AI is one of those promised technologies - it has great promise for the future, but that future never quite arrives. Once again, it is wrong to say that no developments have been made. There have been developments in AI that have formed part of our contemporary every day experience. The use of Chat Bots in customer service might be an example here. However, these developments represent incremental change rather than the radical change originally forecast.
It also gives a clue to how we are likely to experience the greater use of robotics. In the 1950s, the view of how we would use robots today was very much along the lines of domestic servants. What has actually happened is the robots have been incorporated into the appliances that the domestic robots would have been operating. It is the internalisation of automation, robotics, and AI that we can see all around us rather than the creation of some huge army of 'Robbie the Robots'.
There is one further factor that argues against the narrative of the robot overlords - demographics. Western societies are currently ageing. Older workers are leaving the workforce in increasing numbers. This is starting to have an impact on the pool of labour and is set to increase across the course of this decade. It is entirely possible that any increase in automation could be the result of the displacement of labour - from the workforce into retirement - rather than the cause of it. If the currently tight labour markets tighten even further, employers may well experience a cost advantage to invest in automated systems. In this case, the use of robots would be to augment the workforce rather than to replace it.
Which brings us back to the original question. Where are the robot overlords? They are obviously not as evident as forecast ten years ago. The original forecast was flawed because it didn't allow for the long term impact of the global financial crisis. It was also seduced by the promise of potentially disruptive technologies. That promise has not fully materialised. The technology that has been developed has been used to augment existing systems rather than replace them. And the workforce is reducing as part of a natural process in any case.
To many observers, the wild forecasts of job losses from ten years ago did seem a bit extreme. The numbers suggest that caution should have been the order of the day. More jobs were created than lost in the past decade and the future visions of mass unemployment now seem rather fanciful. However, having said that, this view of the future does still creep into the work of some members of the futures community. This is more to serve a political narrative of the future. It is really a question of belief. If someone really wants to believe in robot overlords, no amount of persuasion will help to change their minds.
Stephen Aguilar-Millan
© The European Futures Observatory 2021
No comments:
Post a Comment