There is a number doing the rounds in Australian boardrooms that manages to be encouraging and alarming in the same breath. New modelling from EY-Parthenon suggests artificial intelligence could add as much as $116 billion to the national economy, a figure large enough to reshape budgets, industries and the way we talk about productivity. Sitting inside that optimistic headline, though, is a harder truth: the same technology capable of generating all that value could also perform close to a third of the jobs Australians currently do.
That tension is the starting point for a pointed opinion piece by futurist and work strategist Dom Price, writing for Forbes Australia. His argument is not that the robots are coming, a framing he clearly finds tired, but that Australia is sleepwalking through a transition it can actually see arriving. We have the modelling, we have the warnings, and yet, he contends, we are not planning for the disruption and the shortage at the same time, even though both are unfolding in front of us.
What the modelling actually says
The EY-Parthenon modelling lands squarely in the middle of a debate that has been running hot since generative AI went mainstream. On one side, the technology promises an economic uplift measured in the tens of billions, driven by faster processes, automated administration and knowledge work that no longer needs a human at every step. On the other, the very tasks that make those savings possible are the tasks people are paid to perform today.
Estimates that AI could touch or automate roughly a third of Australian roles are not new in spirit. Similar figures have appeared in work from the Productivity Commission and various consultancies over the past two years, and they tend to cluster around white-collar, process-heavy occupations: administration, data entry, parts of legal and financial services, customer support and middle management. The point Price presses is that these numbers rarely arrive alongside a plan. We treat the $116 billion as a prize to be claimed and the job displacement as weather to be endured.
The shortage nobody mentions in the same sentence
What gives the piece its edge is the second half of the paradox. At the same moment AI threatens to hollow out swathes of desk work, other parts of the Australian economy cannot find enough people. Aged care, disability support, construction, nursing, teaching and the skilled trades have all been flagged as chronically short-staffed, and no chatbot is going to install the wiring, lift the patient or stand in front of the classroom.
Price’s frustration is that these two stories are almost never told together. Policymakers and executives discuss automation in one meeting and workforce shortages in another, as if they were unrelated line items rather than two sides of the same structural shift. The obvious question, and the one he keeps returning to, is why the country is not actively planning to move people, funding and training from the shrinking side of the ledger to the growing one. A displaced accounts clerk is not automatically an aged care worker, but the gap between the two is a policy and training problem, not an act of nature.
Two ways to read the same forecast
There is a genuinely optimistic case here, and it is worth stating plainly. If AI absorbs the repetitive, low-judgment tasks, the argument goes, humans are freed for work that needs empathy, creativity and physical presence, precisely the roles going unfilled. Businesses capture the productivity gain, workers move up the value chain, and the economy grows without leaving people stranded. In this reading, the $116 billion and the labour shortage cancel each other out in the best possible way.
The pessimistic case is just as coherent, and history gives it weight. Transitions of this kind are rarely smooth or evenly distributed. The gains tend to accrue to firms and shareholders quickly, while the costs land on individuals slowly and unevenly, concentrated in particular regions, age brackets and industries. Retraining a mid-career worker is expensive, disruptive and far from guaranteed, and without deliberate intervention the default outcome is not redeployment but unemployment on one side and unmet demand on the other. Price’s central claim is that the difference between these two futures is not the technology. It is whether anyone chooses to plan.
What it means for Australia
For Australia specifically, the stakes are sharpened by the shape of the local economy. We are a services-heavy nation with a large public sector and a comparatively small manufacturing base, which means a great deal of national employment sits in exactly the administrative and knowledge roles most exposed to automation. At the same time, an ageing population is driving relentless demand for care work that is difficult to automate and hard to staff. The mismatch Price describes is not a hypothetical imported from Silicon Valley; it is arguably more acute here than in many comparable economies.
The policy machinery is beginning to stir. The federal government has been consulting on a national AI approach, the Productivity Commission has folded AI into its productivity agenda, and unions have started pushing for a seat at the table on how the technology is deployed. What remains thin is the connective tissue between AI strategy and workforce strategy, the funded, practical programs that would help a person leave a role AI can do and step into one it cannot. Skills funding, portable retraining, regional support and clear signals to employers are the unglamorous work that turns a forecast into a soft landing.
What’s next
The immediate test is whether the coming wave of AI strategy documents and productivity reviews treats displacement and shortage as a single problem or two separate ones. If the modelling is even roughly right, the $116 billion will not simply appear; it will be earned through a reorganisation of work that touches millions of Australians, and the quality of that reorganisation depends on decisions being made now, before the disruption peaks rather than after.
Price’s parting message is deliberately unfatalistic. AI replacing Australian jobs, in his telling, does not have to be a disaster, but avoiding one requires the country to stop admiring the upside and quietly bracing for the downside as if they were disconnected. The tools to plan exist. Whether we use them is, for once, entirely up to us.
Sources: Forbes Australia.
