Human training to rise along with the machines

Working Australians will need to spend an extra three hours a week in education to adapt to future automation in the economy.

AlphaBeta’s Andrew Charlton. Picture: Hollie Adams

Australians are likely to spend an extra three hours a week in education and training across their life, on average, to learn the skills needed to meet the challenge of automation in the economy, accord­ing to a new economic analysis.

The study, Future Skills, commissioned by Google and carried out by economic consultancy AlphaBeta, says people will need, on average, to double the share of learning they do when over the age of 21 to ready themselves for the age of intelligent machines.

Australians get about 19 per cent of their learning (measured by hours spent in education and training) when over the age of 21, but this will need to rise to 41 per cent by 2040 to meet the need to learn new skills.

The report says three forces — increased longevity, automation and less predictable career paths — are behind the push for more lifelong learning.

It found there were three groups of skills needed in the workforce and demand for some would diminish while demand for others would increase as automation grew.

Knowledge skills — such as maths, language or medicine — which are the ability to retain specific information used to perform a work task, increasingly will be the domain of the machine.

Abilities — such as physical strength, mental competence, driving or time management — also are prone to takeover by machines in many instances.

But characteristics — such as creativity, integrity, leadership, persistence and empathy — are in a human domain where machines are least ready to intrude, and people with these skills will be increasingly valuable to employers.

All change in the job market.

However, AlphaBeta director Andrew Charlton said there was still a strong need for people to have broad skills, and not just limit themselves to mastering characteristics, because things such as leadership or empathy alone were not enough to accomplish a task.

“People who say machines can do X so humans don’t need to are wrong,” Dr Charlton said. “Humans will need that skill in combination with other skills that are uniquely human in order to get tasks done.”

But he said it also would be a mistake to try to replicate in detail specific skills that machines were good at, such as coding.

“We’re looking for almost the opposite of that — encouraging people to build skills which enable them to work with machines rather than compete with machines,” he said.

Dr Charlton built his analysis based on a forecast that each Australian worker, on average, would change jobs 2.4 times between now and 2040. As a result of using data about the tasks required in each of Australia’s 348 occupations, he estimated the average change in tasks for each worker across 10 years at 18 per cent. By applying data about the training time required to learn particular skills he was able to calculate the volume of training likely to be needed.

If Dr Charlton is correct, education providers have an opportunity to grow enormously in the next two decades.

Dr Charlton estimated the stock of education (that is, the time already spent learning by all Australians) at 300 billion hours. This would need to rise to 600 billion hours by 2040, he said.

Each person will need 8000 hours more learning, on top of the average of 24,000 hours people currently spend learning across their lifetime, and this will need to be done by adults as there is no room to add more hours to school learning.