Seven years of zero productivity growth: AI is the lever to pull

Former RBA Governor Philip Lowe opened the EO Regional Conference at Port Stephens this week with a clear and sobering assessment of the Australian economy which matches my view of our economy.

His argument, stripped back:

• Average living standards have gone nowhere for seven years, because output per hour worked has gone nowhere.
• From the late 80s to around 2018, productivity grew at 1.5 to 2 per cent a year. Had that continued, Australians would be roughly 10 per cent better off today, and the gap widens every year.
• For context, a one percentage point rate rise takes about one per cent out of demand over a couple of years. The productivity shortfall dwarfs anything the RBA can do.
• He proposed six fixes - tax, IR, business regulation, energy, education and migration - these are all stuck in Canberra with policy stagnation.

Two points stood out. Productivity isn't people working harder; it's giving them different tools, processes and systems.

On AI, Australia talks almost entirely about the risks while the US talks about the upside - I see this inside Aussie businesses every day and it floors me.

My view: waiting for policy reform is not an attractive strategy for business owners. Instead, AI is the cheapest way any business can deepen its capital stock right now, and very few are doing it properly.

Most Australian companies have bought licences, run a few pilots and declared progress. That isn't productivity.

A licence is not a redesigned process. Productivity shows up when you rebuild a workflow end to end, measure the hours removed, and decide deliberately what to do with them.

For CEOs and boards, your actions are clear:

1. Stop measuring AI by adoption. Measure output per hour, per process.
2. Pick the three workflows consuming the most skilled labour and redesign them, rather than bolting AI onto them.
3. Fund AI as capital investment with an ROI case, not an IT line item.
4. Plan for higher rates.

At Propel Ventures we've spent a decade de-risking high-stakes digital products, and our AI enablement and deployment programs exist for exactly this: moving AI from pilot to measurable change in how work gets done.

PS. He is a lovely and intelligent man, it’s a shame our RBA is hamstrung with just one tool (interest rates) when that is relatively impotent and unfair in its application when instead we need real fiscal policy reform.