Former CSIRO chief executive Dr Larry Marshall’s innovation worldview was shaped by Stanford University and Silicon Valley in the US more than his Australian origins.
These days, he’s on the board of mining company Fortescue, chairing its Innovation Committee, also chairs the American Chamber of Commerce in Australia (AmCham) board and is a member of the Australian Government’s Circular Economy Ministerial Advisory Group.
He’s the author of Invention to Innovation: How Scientists Can Drive Our Economy (written with Jenna Daroczy, CSIRO Publishing, 2023).
The Fifth Estate’s Murray Hogarth caught up with Marshall at the UN Global Compact Network Australia two day business sustainability event in Sydney recently.
Q and A
Does Australia really have a net zero plan?
When I was running CSIRO, we did a first cut of a net zero plan for Australia in 2016 and then we fine tuned it by about 2018. It’s more or less the plan that the current government is using to get to net zero, and it shows us achieving net zero by 2050. There are some wicked, hard problems in that plan, though. It showed that you still needed about 6 to 8 per cent gas to help maintain grid stability, versus deploying an inordinate amount of storage to stabilise the grid – and for the other two-thirds of our emissions that don’t come from electricity. And that was the wicked problem that we really focused on. Is there a way to do it that’s still economically viable, and that won’t drive up the cost of energy and industry? So, we have a plan, but, as a nation, we’re probably not doing too well on following the plan, but neither is the world.
So, we know the problems?
I think we do, yeah. I don’t know that the public does, but in the science-engineering community we’re very focused on those challenges.
A 100 per cent renewables-only strategy is not going to work?
That’s not what our strategy is; remember, renewables are for electricity – only one third of our emission challenge today. There’s an event in Kyoto every year called the RD20. It’s the heads of the National Science agencies of the G20 countries. “A” goes first, and when I presented our net zero plan, it showed we had about a 30 per cent gap between what we knew we could achieve in terms of technology and proven technology scales versus what we still had to solve. This was controversial back then, but despite bold political commitments to the contrary, slowly, as the hours went on, around the table, most G20 countries got up and admitted they had a similar gap.
A victory for transparency?
The end outcome of that transparency was that the United States invited us as the first foreign country to join their net zero mission. We’d created a national mission for net zero [Towards Net Zero] in Australia, but they invited us into theirs, which was a huge step. I think because we’re transparent, and I reckon that’s part of our challenge, whether it’s a company or a government, this is really hard, and the more we try to say we’ve got a solution, the more we play the politics, the less we’re telling people what the real problems are. And we’re not going to solve it if we don’t face up to the fact that there are some wicked problems that we need help to solve collaboratively & globally.
Is that actually an optimistic scenario? If we’re talking about 2050, it’s 25 years away; we know the problems, we’ve got a reasonable idea of the solutions, and we know that it’s hard in places.
Let me give you an example. About a third of our emissions as a country come from electricity generation, and they’re easier to abate because solar and wind can take care of them. But there’s a problem in that: in order to get the same stability of electricity that we enjoy today, you have to deploy 100 times more energy storage. That becomes very expensive.
You’ve said we could also redesign the electricity grid.
Australia has the longest grid in the world, but it’s a point-to-point grid. And to do renewables, you need a mesh. You need to be able to move energy seamlessly from one place to the other. When you can do that, you can get stability as well. And actually, you can use AI to anticipate the variation in wind and solar and move the energy around on the mesh to give you stability, almost like Dolby Noise reduction for energy. But you can only do that with a mesh network or a whole bunch of interconnected micro grids. Whichever solution you pick, it’s very expensive because we don’t pay for stability today, we take it for granted.
Which brings us to hydrogen to displace the six to seven per cent gas in the 2050 net zero plan
Hydrogen could be a game changer. It has the ability to replace gas in the electricity application and in the larger industrial applications that depend on chemical energy, not electricity. Unfortunately, it’s more expensive than gas or diesel, but so is deploying ten times more battery capacity converting the grid to a mesh or burning diesel to contaminate the planet. Hydrogen is a good alternative, and that’s why we backed it so heavily. Fortescue made a big breakthrough in building and sailing the world’s first ammonia powered ship – green ammonia comes from hydrogen instead of gas, and competes directly with oil & diesel, and is real zero, not just net zero.
So that’s quite a very targeted intervention, with the heavy lifting being done by lower-cost technologies
Exactly. Another of the wicked problems we talked about in 2018 was agriculture, which is about 15 per cent of global emissions. But if you feed seaweed to cattle, you pretty much eliminate their emissions. Aussie companies like Sea Forrest are growing this product, using agriculture itself to solve agriculture’s emissions, showing you can have economic growth with emissions reduction. So, the fact that we could do that in Australia, solve a global problem of 10 to 15 per cent of the world’s emissions with science, gives me great hope – also real zero.
But the last bit, the wicked problems amounting to about 20 per cent of the net zero by 2050 task, are still there?
Even we, as the national science agency (CSIRO), with an emissions profile like Australia’s, across our 55 sites nationally; we only got 83 per cent of the way to net zero. We couldn’t crack that last 20 per cent, and that’s what all countries are struggling with. It’s wicked hard, even though they might have quite different scenarios and ways of getting there. The US, for example, and Europe, because they have a very established nuclear industry, they’re looking to nuclear (fission) to solve some of their wicked, hard challenges because they’ve got that history which we don’t. A nuclear submarine engine could power a data centre, and some tech giants are pursuing that to eliminate emissions & insatiable power hunger of AI. The US is also looking at nuclear fusion, which is an amazing technology but is still being invented.
Back to your earlier point, Australia has a plan, but we’re not necessarily following it that well
Look at the government’s target setting and the priorities that they’re making, and things ARENA is doing, and things that the Clean Energy Finance Corporation is backing, and the Future Made in Australia plan, and that is all pretty much what the CSIRO started doing five, six years ago. But remember, the government hasn’t moved very fast in terms of reducing its own emissions, and neither has the world.
Let’s turn to AI. Can it help get us out of these wicked problems, or will it create new ones?
At CSIRO, we deployed AI back in about 2017, but not true AI in the sense that it runs by itself. [it’s] Not a black box; you can’t interrogate it; it just spits out an answer. We never did it like that.
We did the machine learning and predictive analytics algorithms, the basis of AI, but we always had humans steering it and engaging it and kind of making sure that we understood what it was doing. Take getting to a higher-yield wheat despite climate change. That’s a 7 to 10 year breeding cycle to breed a strain of wheat that won’t suffer under climate change scenarios. AI can accelerate that down to about maybe two to three years. So, it’s a powerful tool. It’s a great lever. But what we experienced when we tried it, the geneticists in CSIRO and the agricultural scientists, they’ve got an instinct for what’s going to work. So, the AI just tries every possible scenario, billions of combinations it doesn’t actually understand, and it uses the data to feed it and steer it. When you put a human in the loop, who’s an expert, and steer the AI, you get a far better result than the AI could do on its own. And we found the same thing with bushfires. Google did a project in Google X trying to make an AI bushfire simulator using 20 years of images from Google Maps, essentially learning from the past but without the science.
We made one way better in Australia by using Western physics and Indigenous experience and knowledge of bushfires in a human-AI collaboration. So, I think that’s a better future for us, AI as a collaborator. I’m not that comfortable with the future where we just let AI do everything and trust it.
You are also chair of AMCHAM. How does the US-Australia relationship play out now?
I think that relationship’s rock solid; it’s based on values, business, investment and shared history. I don’t think we need to worry about that, no matter who’s in charge.
What about the Biden Administration’s big climate action and low emission technology funding initiative, the Inflation Reduction Act (IRA)?
Will the IRA stay the same? I reckon the Republicans might rebrand and tweak it a bit. But the IRA is important for America and Australia. Like Canada, we get special status for the IRA. And now, globally, it’s hard to find places where energy is cheap enough to make green products like hydrogen or ammonia.
The IRA artificially lowers the cost of energy by about 30 per cent in the US for a Hydrogen project. That’s why Fortescue went to the US to do bigger-scale hydrogen rather than doing it in Australia because the energy costs are too high here. There’s a risk that Trump might tweak that, but I don’t think very much because it’s bringing so much business to the United States, and he’s very business focused. People worry a lot about, you know, commitment to emissions and all the rest of it. When you look at the numbers, US emissions went down a little bit more in Trump’s term than under Biden.
Do you see space technology as a US-Australia opportunity?
Trump was really big on space and Australia. Our government backed off on space a bit in Australia. If space becomes a bigger priority again, I think that’ll be good because there are a lot of space startups in Australia; we really do have a lot of capability there. The other one is for the net zero transition. There are some key bits of technology, like lithium. The US is worried about China’s role in supplying those. I think there’s a chance for Australia to jump in and earn our way a little bit further up the value chain, and for the US to encourage that, I think they feel pretty comfortable with us making higher value components to feed their value chain.
Has it turned out to be an economic blessing that Australia stopped making cars?
We can’t compete on the cost of labour if we make commodity products, and countries like China are so well set up. Back in 2011, Alan Finkel, Evan Thornleigh & I tried to persuade the government to bet on EVs made in Australia – a second chance but on something unique instead of a commodity.
I think where we perform really well is making things from very unique materials. So I could see Australia making, for example, core components in lithium batteries, not the whole battery, but the critical elements, because we control the raw material, and we have the materials science. Australia is really good at materials science, and I think we have the ability to create those critical components that go into everybody’s batteries rather than trying to build the whole battery itself.
Similarly, with green steel, if we can make green iron and embody renewable energy and get rid of the emissions, do all that all year, we don’t need to make steel, and we can buy the cars back from China because they’re made with Australian green iron and Australian unique materials.
The other one is the Chips Act; the US recognises they’ve fallen behind in semiconductors. Turns out Australia has some amazing semiconductor capability, but it’s pretty much all in the R and D space. Australia’s got a chance to jump into that supply chain as well. It’s actually a second chance for us to jump into that as the US tries to rebuild it.
Will we come under more pressure to take American technologies, including nuclear?
Historically, the US hasn’t had any trouble selling its technologies around the world. I don’t think we’ll see that, but, but I guess the thing I worry about more is China. I’d hate to see a situation where Australia had to choose between the two.
I think governments in Australia, or both sides, have done a good job of positioning Australia as an honest broker in the middle. The US has never asked us not to sell our iron ore rare earths to China, but they are asking us to help them rebuild their supply chain, and that’s a fantastic opportunity for us. I don’t think China is going to get too upset about that because the need for these semiconductors is growing exponentially like AI is. So, I think there’s enough we can supply both. There’s room for everybody to play. It probably reduces tensions a little bit.
Your book Invention to Innovation envisages Australia being more like America and Silicon Valley, not just inventing stuff but creating innovation businesses in much larger numbers. Is the net zero transition a prime opportunity for that?
It’s absolutely perfect. And we’re more at the forefront of it than people realise, because we’ve got the biggest penetration of solar in the world. We actually invented the cell design that’s used around the world. We just didn’t turn it into a product. Wish we’d done that, but we sort of got a second bite of cherry here, though. And there’s a government pushing a big circularity agenda, and solar panels and lithium batteries weren’t necessarily designed to be easily recycled, and so Australia’s got a chance to have another crack at solar and batteries, to reimagine them as recyclable products. And that may be our last shot at those two technologies because otherwise, China owns it, and frankly, they earned it by investing massively at a time when others refused to invest. But we have a second chance, let’s not waste it!
