If you are a senior leader in Australian Heavy Industry, the headline number I want you to sit with for a moment is not the one you've seen in the trade press about the AI bubble, nor the one about white collar jobs disappearing. It's this:

Worth of AI infrastructure investment announced for Australia in 2023 — already the largest technology build-out in this country's history, and it has barely started.

The AI data centre build-out is already creating wage escalation that should concern every project manager and commercial manager in Heavy Industry. Electricians specialising in high-density data centre infrastructure — 480V busway systems, precision power distribution — are today commanding AU$120,000 to AU$150,000 per annum base in the Australian market. Commissioning specialists are being locked into projects 12 to 18 months in advance. These are not future projections. This is the market right now.

The structural reason for this pressure is not difficult to find. Australia's national data centre capacity is projected to grow from approximately 0.3 GW in 2024 to between 2.2 GW and 3.2 GW by 2035. That is a tenfold increase in a decade. The data centre construction market alone is forecast to grow from AU$3.64 billion in 2025 to AU$6.24 billion by 2034. Every megawatt of that buildout requires electricians, HVAC specialists, structural trades, and control systems technicians.

Here is where I want to be direct with you, because the figures circulating in the general media have significantly understated the situation. Australia does not face a shortage of 17,400 electrical workers by 2030, as some earlier estimates suggested. The actual validated position, drawn from RACE for 2030 and the Energy Council of Australia, is substantially more severe:

To close even the 2030 gap, Australia needed 20,500 new apprentice electricians to commence each year between 2024 and 2030 — a rate 40% above the historical average. As of today, there is an existing shortfall of 22,000 apprentices just to meet what was already planned. This pipeline cannot be turned on quickly, and will grow until addressed adequately.

For executives in Oil & Gas, Energy, and industrial construction, this matters directly. The same electricians, instrumentation technicians, and control systems tradespeople that power your offshore platforms and process facilities are now being courted at premium rates by data centre developers who can offer fly-in-fly-out schedules (which Oil & Gas & Energy have fought hard to eliminate to control costs) and controlled environments with no weather risk.

It would be a mistake to view the data centre buildout in isolation. Multiple simultaneous demand drivers are converging on the same constrained trade workforce:

The oil and gas sector in Australia already faces its own distinct workforce challenge that predates the AI infrastructure surge. The industry is losing early and mid-career professionals faster than it can replace them. Offshore technicians, mechanical fitters, riggers, drillers, process operators with LNG experience, and subsea specialists are all in record demand and in structural short supply. The aging of the existing workforce is compressing the available talent pool from both ends simultaneously: fewer entrants from below, accelerating retirements from above.

There is a nuanced counter-argument worth acknowledging, and it is one that a serious executive audience deserves to hear. Automation is real. Logistics, basic assembly work, some predictable maintenance tasks, and certain operating roles are being progressively automated in ways that will reduce headcount in some parts of the blue collar workforce. The benefits of the AI infrastructure boom will not be distributed evenly across all trade categories.

The executives I work with across Heavy Industry in Australia are beginning to understand that workforce planning is no longer an HR function. It is a strategy and bid management function, a contract delivery function, and a business development function. If you cannot guarantee trades capacity when you tender for a contract, you cannot price it accurately. If you cannot price it accurately, you should not be bidding it.

The organisations that recognise this and begin building their trade workforce capacity and their AI augmentation capabilities in parallel are the ones that will define the competitive landscape in Australian Heavy Industry over the next decade.

At AI That Works, I work specifically with Heavy Industry contractors to develop strategies that address both sides of this equation — understanding where AI can genuinely augment your operations, and how to build the workforce and commercial capability to capitalise on the coming wave of infrastructure demand.