The role of practitioners
On September 22 Industry Minister Tim Ayres spoke at the UTS launch of Turning AI into Productivity by Emeritus Professor Roy Green and Dr John Howard in partnership with Google Australia. All three voices landed on one important finding; while the technology is becoming available everywhere, the specific factors needed to create value from it are not. The report names what innovation policy has overlooked for decades, translation capability, recognising Industrial AI will not deliver productivity at scale for Australia without it.
This presents us with a challenge which the report highlights; who builds the translation engine, and what they need to do it.
Been here before
Australia has been presented opportunities like this before, but as a nation we have been unable to capture the full potential. The most obvious we explored in what the mining boom did and did not deliver for the Australian economy. In this case while government (rightly) funded the research supply side through collaborative industry funding, policy overlooked the translation engine that needed support to be built in parallel. While commodity rents flowed through our economy, the compounding knowledge rents created from local innovation flowed to overseas companies who recognised the value, and invested in them. The income was realised, but the capability was not retained.
What this report confirms
This is why Green and Howard’s report is so important. It is unambiguous in its examination of 110 innovation ecosystems across 35 countries, and the findings are clear about what is needed. The productive impact of AI concentrates in places with a genuine institutional focus on research translation and enterprise capability. Without translation, the absorptive potential of compute capacity goes unused, testbeds end up being demonstration labs, and the skills required to deliver drain away. This is precisely what happens here in Australia.
Australia is not short of evidence, what is missing is the investment in the places where translation happens, and the practitioners with the commercial skills and experience to build them.
Design is not construct
We need translation capability with defined long term funding mandates and accountability metrics. Innovation policy has consistently consulted with distinguished academics, industry governance experts, and representative associations with strong networks but limited commercial experience. They are the right architects, but design and construction are two separate capabilities, developed through entirely different sets of experiences. The construction phase requires commercial leadership. It requires experience in building, operating and maintaining a commercial enterprise. It requires comfort with accountability frameworks that measure real outcomes; customers, orders, revenue and growth.
Absorptive capacity required
The translation of innovation comes from sustained market engagement, by firms with the practical experience of delivering products and services to address market needs. It requires both technology readiness AND commercial maturity. Success is measured in outcomes not activities. It is revenues generated, number of employees, markets served and sales numbers. Companies succeed through perseverance and accountability. It can be ugly, it can get messy, but as we saw in the Industry Growth Centres when you provide local companies the right support to survive and thrive in competitive markets, you produce the absorptive capacity needed within the local economy.
Green and Howard use the analogy of electricity to explain how general purpose technologies reach economic impact. In this example, impact took more than forty years to materialise, and it was only through the deliberate construction of wholesale industrial systems that benefits were realised. This was not a task of governance, it was a commercial development process, carried out company by company, product by product, customer by customer. As the Minister said in his speech at the launch, value is not generated by the technology itself, but what those with the expertise do when they apply it to real industry needs.
Two instruments, not one.
The pipeline of research is one critical asset that Australia has built and must maintain. It is the sustained investment into university research, into co-operative research centres, and into programs like the R&D Tax Incentive to support a world-class supply of innovation. It is a foundation which the future opportunities like Industrial AI, or Quantum, or clean energy depend on. But the foundation is not the building.
The missing instrument is support for local Australian companies positioned to translate Australian research into competitive products and services. They are the established vendors and suppliers with market access, customer relationships and the operational experience to take those products to market. The evidence this model works is not theoretical, as we discussed in the missing middle. And while many of our publicly owned industrial companies were lost to privatisation decades ago, there are still many local firms that do have the requisite skills and capabilities to scale.
It is not that we do not have the capability, it exists but is overlooked. As a consequence it ends up underfunded, under supported and risk disappearing through acquisition and attrition. Local companies need the right incentives and support, so that when our Industrial AI policies arrive, our economy has the absorptive capacity to convert them into long term benefits for the country.
Time to build
Turning AI into productivity has opened up the right conversation that has been needed for years, and all the right signals are there. We need sustained investment in both the supply side and the delivery side of innovation, with practitioners from both sides of the value chain involved.
The question is how do we accelerate the construction phase, led by those with the experience and proven track record of delivering measurable outcomes necessary for translation to succeed.