The debate over artificial intelligence and its economic impact has intensified, with economist Greg Jericho raising pointed questions about what Australians will gain from the current investment boom. Writing in The Guardian, Jericho, a self-described sceptic of AI productivity booms, argues that the data so far points to a future dominated by energy-intensive data centres that contribute little to broad economic prosperity.
Jericho's analysis comes amid a flurry of announcements from tech giants and governments alike touting AI as the next industrial revolution. Australia has seen a rush of proposals for massive data centre projects, fueled by global demand for AI processing power. However, Jericho suggests that much of this investment is being directed toward infrastructure that may not yield the productivity gains promised by AI advocates.
"Booms in investment usually lead to economic growth," Jericho writes. "But this one looks unlikely to translate into better living standards for Australians." He likens the current fervour to past technology bubbles, warning that the primary beneficiaries may be a handful of large technology companies rather than the broader population.
The commentary reflects a growing chorus of economists and critics who question whether AI will deliver on its lofty promises. While proponents argue that AI will revolutionise industries from healthcare to finance, sceptics point to the immense energy consumption of data centres and the lack of clear, widespread economic returns so far.
Jericho's piece does not dismiss all potential benefits but urges caution, particularly as governments consider taxpayer subsidies for data centre projects. His analysis is a counterpoint to the optimistic narratives often promoted by industry leaders and policymakers who are keen to position Australia as a regional AI hub.
As the debate continues, the question remains: will the AI boom translate into tangible improvements in productivity and living standards, or will it result in a costly overinvestment in infrastructure with limited societal payoff?