Artificial intelligence is often described as software that lives in the cloud. In reality, the cloud depends on physical infrastructure: data centres, servers, chips, cooling systems, electricity networks and skilled workers.
Reserve Bank Governor Michelle Bullock has recently identified the AI and data-centre investment boom as one of the factors adding to demand and potential inflationary pressure. The concern is not that technology investment is inherently bad. It is that a rapid increase in demand for energy, equipment, land and construction can compete for resources when supply is already constrained.
For Australia, the question is how to attract investment while ensuring infrastructure can support it. Data centres can create construction work, technical roles and demand for local services. They can also require substantial electricity and water, depending on their design and location.
What should be assessed?
Communities and planners should ask how a proposed facility will be powered, what grid upgrades are required, how cooling will work and whether the project competes with other local needs. Businesses should consider whether the cost of AI services will remain predictable as computing demand grows.
Government policy needs to balance digital competitiveness with reliable energy, environmental standards and transparent planning. A data-centre announcement is not proof that the benefits will automatically flow to local households or workers.
The AI economy is built on infrastructure. Understanding that physical foundation helps explain why the technology debate is also an energy, planning and economic policy debate.
Read more: ABC News: RBA comments on AI and data centres; Reserve Bank of Australia.
