AI could shatter Australia’s labour market

Experts warn that half of the entry -level jobs may disappear. Australia ignores an artificial intelligence crisis that can rob economic leverage workers. Jemma Nott Reports.

Many former and existing AI researchers or AI heads warned that the US labor market may face significant deductions from AI.

Darios AmodeiCEO AnthropicHe went so far to say that employee “Economic leverage may lose.” The workers warned that all entry -levels may lose half of white -collar jobs and that unemployment may rise to 20% or 30% as well as losing the general loss of the ability to shape the direction of the labor force. Among the economic analysts of the labor market, there was a general attitude that accepted legitimate concerns about AI, but deprived of imagination about a world where people like Darios could really be real.

This suspicion of narrative, sold by various experts in the technology industry, is not affiliated to a large technology company, but is common among Australian analysts and political economists, because these companies are interested in panic to invest. However, what is not considered in detail is that it may be true that both those who are suspected of strange claims made by technology giants and those who think that we are 5 years away from a dystopic artificial general intelligence.

Undoubtedly, in the context of AI, a bubble is formed with most of the features of most of the 90s point-com balloon. The first ten AI company is even more over -valued Just before the DOT-COM collapse collapses, it is more than the best Dot-com companies. Microsoft, for example backward It referred to the lack of economic growth produced by AI from a center of a data center he built. In the meantime, an enormous amount of investment in data centers is shed in the expectation that extremely higher calculation levels will lead to economic transformative. Most of the first ten technology companies, especially, went to rise and entered the Saudi rooms billionaires And Trump ManagementAnd the trillion reached the dollar in the process “too big to fail” situation.

What we can learn from the Dot-Com the bubble accident is that you can make it meaningful by the market to understand extremely worthless companies in a meaningful way (companies have become the key to the “Agent Ai”, which literally adds .com in their names during the DOT-Com the balloon). As a result, the Dot-Com the balloon explosion was not a failed bet that websites would have an incredible power to generate income; It was a failed bet on which websites to create revenues And the risk capitalists of the time did not understand that website traffic was not the same as earnings, but also that experts in Silicon Valley do not correctly understand that productivity alone does not lead to economic growth. In the West, especially with services and financial industries, automation in the labor market will cause the greatest damage in the main driving force of economic growth – consumer expenditures.

However, this does not mean that extreme valuable companies will not radically transform our labor market, and in fact, why these technology headings can make a sharp turn. fascination They themselves to the US government. We keep our heads in the sand in Australia. report with large language models or changed GPT4 and now by GPT5, claiming to be an even better agent AI engineer promise To make software from end to end. The detoctors say they will immediately create new jobs in the form of engineering experts, but what they ignore requires less people and less specialization in every new model that hit the market.

Team by AI 2027 This year, only an unreliable coding agent foresees, but not the rise of the “vibe coding ver by the general laziness, but with the predictions of the past and in fact a conservative rhythm. The middle scene of their predictions left us so smart that they can do almost anything in the degree of computer science until the middle of 2026 and in super intelligence until 2027.

We do not have to believe that radical gains will be made because we work close to Özyinel. self -improvement (Machines Learning to improve themselves) consist of data centers to greatly increase the calculation centers to make the future look scary. The most powerful detoters of this argument seem to ignore that it has more automating ability than launched in the Australian market. McKinsey, 9% of Australia labor It may be necessary to “go through its current roles until 2030.” However, let’s learn more about exactly how it is. We have already automated the large parts of some of the most productive industries such as agriculture and mining in this country, but more of the tasks in these industries can be automated. For example, there is technology expand Broadacre/ Robot fleets in orchards, using mine -diameter Timing representativesAutomatizing parts of the timing producingRobotic Order/ Verification structureend -to -end autonomous supply In Wholesale Trade, Real Time Pricing/Promotion Representatives For retail, for automatic kitchen orchestration In larger food chains and in the white collar world, the very agent is in the first draft outputs (or in other words tasks Completed by a group of AI agents working together).

Some new jobs can be created in artificial intelligence, and the investment in technology will be a pleasant contribution to an economy with less general complexity than Uganda or Armenia, such as Australia. However, the problem only leaves us with incredibly conservative estimates, the next five years when things are quite replaced, and even then, we must assume that a middle -aged truck driver on a mining site will take a job in AI operations without complete government support. Furthermore, we ignore that all the plan, which is still transparent for everyone to see, is a future and a future that does not have a future in which they dominate the business and capital.

Therefore, we cannot allow the productivity commission to ignore that productivity in an economy should not be a last goal in itself. While human labor becomes difficult, productivity increases and displaced, ultimately means not only for anything, not only for the welfare of the citizens of a nation, but also for potentially economic growth. So, ultimately, even though Darios’s concerns only happened to a company of 60 billion dollars resources It is still true for new data centers. We are looking at the barrel of a technology that has the ability to completely remove the traditional economic leverage of an Australian worker. Don’t make mistakes – the truth that we cannot attract our labor against a machine will become a real crisis in the West. In the Australian labor force, as union participation has decreased rapidly since the 80s, we got rid of the historical gains of the trade union movement, so the potential of the modern labor may not be aware of a tragedy. Whether it’s in 2027 or 2050, I hope you need to go there until we find out.

https://www.youtube.com/watch?v=e85auxyw0qyk

Jemma Nott is a post -graduate student and free writer at the University of Sydney.

Support independent journalism subscribe to IA.