Will machines ever replace people and would that lead to mass joblessness?

AI takes automation a large step further. It makes predictions based on large and complicated amounts of data (which humans can do, although more slowly than computers) that are then used to create goods and services beyond human capabilities, such as searching the web for a story, picture or person, and to create whole new process chains and industries based on technologies we’ve never seen before, such as blockchain.

AI brings four primary effects to any economy:

• The displacement effect. New technologies can lead to a substitution of jobs and tasks currently performed by workers. Job displacement has the loudest volume in the debate about AI. And that’s reasonable. The fear of lost jobs is real, and it is and will continue to occur, as it has occurred throughout history when technology overtook human endeavour, most specifically in the 18th century with the advent of the Industrial Revolution. In Australia, in particular, we still suffer nightmares from the vast displacement of jobs in the textile industry during the 1990s. Even though that displacement was caused by tariff reductions and the quest for low production costs— not by technological change—the fear of losing jobs is still high on our list of worries as a nation. Even putting this history into its appropriate historical context, the displacement effect is still not an absolute. Even with the closure of both the textile and car-manufacturing industries, Australia’s GDP and standard of living have still risen. How come?
• The skill complementarity effect. There’s a complementary increase in jobs and tasks necessary to use, supervise and grow from newly created technologies. At the turn of the 19th century, horses and buggies were the mode of efficient and quick transport. When steam engines and cars took over, farriers and blacksmiths—once part of dominant industries—were now in tiny craft-based occupations. Now look at the jobs involved in producing and using cars and locomotives and project forward to driverless vehicles: the requisite number of jobs to support autonomous vehicle software, manufacturing and supply chains will be enormous. The classic case for the skill complementarity effect is the switch in Germany from coalmining to coal machine manufacturing, which in turn became a turbo-boost to the development of a machining-led export-oriented economy. When cars are made in Germany, they’re certainly not made with low costs of labour—a lesson Australia never fully integrated.
• The productivity effect. Lower prices and higher disposable incomes drive increases in consumer demand. Look at the burgeoning number of cafes, fitness centres and speciality retail stores, which has ballooned in the past decade, and the rise in health-care, social-assistance, aged-care and infant-care jobs. Those sectors have experienced double-digit growth rates over the past decade.
• The low cost of capital effect. AI is increasingly embedded in newer and cheaper products accessible to more people than ever before. Have you used apps on your smartphone today? Got an idea for a small business and know it won’t cost much to get an app rolled out? The steep fall in the cost of technology is driving increased access to technology and enlarging possibilities for innovation, in turn driving new forms of higher productivity. This is particularly beneficial for some economic sectors, such as agriculture, in which the application of technology has delivered enormous increases in crop yields at much lower costs of production.

What can Australia do to maximise the opportunities of AI and ward off threats of joblessness?

• Invest in technology for existing businesses, especially manufacturing.
• Ensure equal access to digital infrastructure, especially in a Covid-19-normal world in which access to online services, work and health is now crucial.
• Boost support for new AI businesses.
• Recapture old markets, especially for manufactured goods that new technology can produce more efficiently.
• Increase self-sufficiency in national security.
• Educate and reskill at all levels of society.
• Invest better, smarter and more heavily in AI R&D.

The workforce impact of AI is akin to the impact of computers and the internet. Most future jobs will have an AI component. Some jobs will disappear, and new jobs will be created. Eventually, it will be unimaginable to work without AI, much like it is today to work without a computer or smartphone. The challenge is that the jobs that will be created will be software- and service-oriented, and online. This means that those jobs will be located where the technology is, not where the customer is. This is a fundamental shift in how the economy is being structured in a global online world. Facebook, for example, employs more than 58,000 people but has only 140 employees in Australia.

It also means that foreign companies will increasingly use Australian data to extract income overseas. The question for Australia is whether we want to actively participate in this new global information-based and AI-driven market, or succumb to it.

This article is an extract from Artificial intelligence: your questions answered, a report published in partnership with the Australian Strategic Policy Institute (ASPI).