A new report released by the CSIRO predicts that 44% of jobs currently performed by Australian workers are at risk of automation, especially routine jobs within the services industry such as in the financial sector.
The CSIRO report, "Tomorrow's Digitally Enhanced Workforce", adds to a recent MacKinsey and Company report that predicted job losses of 25% in the insurance sector due to future automation.
But the CSIRO also warned that while technology will decrease some jobs it will also create an increased demand for others.
The report cites the example of bank tellers whose numbers declined by 50% from 100,000 to 50,000 between 1995 and 2015 due to automation. But by making the banking industry more productive, technology created opportunities to sell a greater range of services. This is reflected in the number of finance professionals increasing three-fold over the same period from 32,000 to 90,000.
The report warned, however, that technology has been able to substitute for workers in a range of routine tasks, which has displaced some occupations entirely, particularly in the middle of the skills and income distribution resulting in 'middle class' jobs becoming scarcer especially in the US with profound economic and social implications.
But it also says: "Those workers who have skills which are complementary to technology become more productive (and thus command better wages and job opportunities) while those without such skills, or who perform routine tasks, do worse, as they can be replaced by technology."
The report noted that jobs most susceptible to automation are those that can be considered routine ie they follow well-defined procedures which are suitable for codifying, eg record-keeping and repetitive customer service.
"Routine [tasks are] not the same as mundane [tasks]: there are many mundane tasks that are likely to defy automation for decades to come, eg medical diagnosis and legal writing," it said.
According to the report, cognitive professions such as scientists are at surprisingly high risk of automation. By extension, routine analysis tasks like those undertaken by financial and investment professionals could also be automated especially if Artificial Intelligence applications continue to develop at their current rate.