“The GDP impact here measures in tens of trillions of dollars over the next 20 years,” he said.
“But this opportunity is perishable for the country, there’s urgency, we need to move now.
“I believe that Australian industry is not adopting technology on the whole as fast as it needs to, particularly in the small and medium or mid cap market, and one area that we see that manifest is the decrease in R&D expenditure.
“We’re roughly 1.9 per cent of GDP versus others that are at 3.5 or 3.6 per cent of GDP.
“As we go through these shifts economically, every industry becomes more data driven, and machine learning and AI sits at the centre and is the basis of competitive advantage.
“We need to lean into this, and we also need to create sovereign capability. We can’t import that values-based system for ethics, we need sovereign AI capability.”
Mr Turner said CSIRO is investing $38 million in AI, together with university partners. Ethics and ethical machine learning will be a stream within that.
“We think it’s a $315 billion GDP opportunity to increase our use of ethical machine-learning based systems and apply those to our strength industries whether its financial services, energy, mining,” he said.
“Over the next decade that’s the prize, $315 billion of GDP impact. And if we get it wrong what we’ll see is global platform companies come in and dominate our industry sectors and extract the value out of our value chain.
“We’re optimistic, we think all the ingredients are here, we think ethics and the way we approach and deal with ethics is at the core and the opportunity to build new tools could put Australia in a really strong place in the global context.”
EY Director, Christina Larkin also spoke at the CEDA event and said AI adoption has been strong across all sectors in Australia.
“We’ve seen a lot of clients start to apply AI in their business, the health sector in particular has been very good,” she said.
“We have a client for example who is applying it in breast cancer research and identifying breast cancers. Another client is looking at the application of AI in IVF to identify which embryo is the best to implant in a procedure. In the energy sector we’ve seen some really great applications in optimising energy models.
“In every sector we’re seeing the application of AI. Even some of the most archaic sectors, the ones that haven’t been the best at adopting technology in the past, they’re really latching on to the opportunity that we’re seeing.”
Standards Australia Policy Manager, Dr Jed Horner said the challenge with AI in the standards world is that it tends to be very technical, but that’s not a bad thing.
“If I use an analogy from the construction sector, they’re very prescriptive and that’s great. It’s around the quality of concrete and all the constitutive elements and you can use that analogy with digital infrastructure too,” he said.
“And I think we need to get to those specifics. So, if we’re talking about privacy, define what the actual issues are, not just a broad notion of ‘I’m worried about my privacy’, that doesn’t mean anything, and in fact there’s some massive inhibitors to AI through our privacy framework in Australia.”
Dr Horner said the international community was already providing a good starting point on how to standardise AI in Australia.
“The OECD have released principles around AI which have been signed up to by our government with 41 others,” he said.
“The hard work will begin in the standards process proper; committees comprised of corporate interests, of government, of academia, of civil society, they all have to be around the table.
“But whatever we do has to be commercially realisable, and real, and it also has to resonate with the broader community. So, it’s striking a point of consensus which is not always what we see with either purely government driven processes, or purely market driven processes.”
Microsoft Corporate, External and Legal Affairs Director, Tom Daemen and Allens Partner, Gavin Smith, also spoke on the CEDA panel. Audio from the event is available via the links below.
Adrian Turner, Data61 MP3
Moderated discussion MP3