Over the last few years, there’s been a common theme running through many discussions about automation that focuses on how many jobs will be lost. But that view is skewed by a number of factors and it grossly understates what humans bring to their work and how intelligent automation really is. Automation may help to do things right, but we need people to ensure it does the right things.
In 2013, Professor Michael Osborne and Dr Carl Benedikt Frey from the The Oxford Martin School at the University of Oxford released research into the future of jobs in the United States. A single ‘fact’ was distilled from their research: almost half of jobs in the US “are at risk of automation”. Since then, a number of other studies have been published pointing to vastly different outcomes.
The problem is that while many tasks can be automated, jobs are much more than just tasks. It takes human intelligence to know what tasks to perform, what order to perform them, how to deal with unexpected challenges and how to be creative.
Osbornbe and Frey noted five years after their initial paper was published that their study, and almost all those that preceded and followed it, omitted one crucial detail: time. No one could accurately predict how long it might take for software to automate routine tasks. Nor did the studies consider that as people were freed from repeatable tasks, they could learn new skills and be engaged in other activities.
Sign up for SmartCompany newsletter.
Free to your inbox every weekday
So, while the headline painted a scary picture, it was a superficial reading of a study that painted an incomplete picture. Indeed, many other researchers, using the same pool of data used in the Oxford study, found that perhaps just 14% of jobs were at risk. The difference in the outcomes from the same pool of data stems from the analysis. Osbornbe and Frey rated jobs with either a 1 or 0 depending on if the job could “be performed by state of the art computer-controlled equipment”. It then relied on what the researchers found were “the occupations about which we were most confident”. And that led to a count of just 70 jobs.
If we think about the millions of different jobs people do, 70 is a small sample upon which to predict the future of all jobs. It’s also important to note that many jobs consist of many different tasks that may be automatable to different degrees.
The sort of revolution that automation will bring over time to workplaces is not new. Electricity and the internal combustion engines made thousands of jobs redundant a little over a century ago. Australia employs about 2% of its workforce in agriculture, down from almost half the workforce at the start of the 20th century. Yet, our national unemployment rate has remained steady at around 5% for many years despite the disruptions brought by computerisation, automation and other technologies.
Workplace revolutions that leverage new technology do not happen overnight. And while there is disruption, people do learn new skills and take up new jobs. There is nothing to suggest that increased use of automation in workplaces won’t result in the same thing.
With workplace automation, using software robots does require a level of intelligence to automate tasks within departments such as AP, AR, sales, operation, etc, or as application bridges (collecting, moving and collating data between discrete applications). The software robot needs to ‘know’ where the data is and how to perform some sort of calculation or manipulation and account for known problem types. But if we haven’t accounted for something unexpected, its lack of inherent “human” intelligence becomes obvious. And this is why people will always be needed.
When faced with an unexpected situation, humans remain far more capable of quickly assimilating new information, adapting and in some cases, updating automation workflows with the new information. When automation frees up time by relieving us of the day to day process tasks, we are able to use some of that time to enhance our training in emerging areas and use our wisdom, creativity, collaboration, problem solving and insight to see new opportunities and ways of doing things.
Indeed, research published by The Economist, suggests that some jobs, even in industries that have seen significant automation, are hard to displace as workers have found ways to maintain productivity in both routine tasks and social and creative ones — something computers or robots cannot do.
There’s no doubt that automation will change how we work. We’ve seen that repeated countless times throughout human history. But humans have never stopped working.
We may do different things but even the most intelligent automation requires humans to create the algorithms, to manage the exceptions, and solve new problems in creative ways.