Compute in Australia: Announcing my new podcast series

Hey everyone, this is a quick announcement to let you know that I'm running a new series of episodes on the podcast.

The series is called 'Compute in Australia', and I'll be exploring the question of whether and how we should scale up a large domestic compute industry here. Think many gigawatts of new AI data centres, some potentially drawing a full nuclear reactor’s worth of power or more – an order of magnitude larger than any clusters currently operating on the continent.*

Compute simply refers to the processing power produced by special chips sitting in racks of servers which are stacked inside those data centres. It's become the key input for both the training and running of modern AI systems.

AI will be at least as big as the internet, maybe as big as the Industrial Revolution. And like past tech transitions, it’s likely to shake up the global distribution of wealth and power. So what is a middle power like Australia to do? What would it look like if we were extremely proactive about ensuring that we come out of this transition better than we entered it? Where in the AI value chain can we become a significant player? 

At least to begin with, an obvious answer seems to be: compute.

I've been ruminating on this for a while now. Early last year, I started thinking about the economic opportunity – whether or not compute could become our next great export industry – and asking a bunch of economist friends what the answer to that question might be. But this year, I started thinking a lot more about the strategic leverage compute could afford us.

That new thinking was primarily motivated by my growing anxiety about our access to frontier AI. If the connection here isn’t immediately obvious, let me explain.

Pretty soon, AI will be so deeply woven into the substrate of society that it becomes a utility like water or the internet. It’s hard to picture this possibility today, especially if you view AI as nothing more than frivolous chatbots. But imagine a not-too-distant future where almost all economic and social activity – from how businesses negotiate and transact to how you access hospitals and government services – is in some way mediated through AI agents.

At that point, uninterrupted access to the AI substrate becomes critical. To take the internet as an analogy, Australia – more precisely, the University of Melbourne – first gained access to the global internet in 1989. If you'd switched off our access in the early 90s, you would've pissed off some university researchers and early adopters, but you wouldn't have triggered societal upheaval. If you switched it off today, the story would be quite different. 

Of course, today’s internet can’t actually be switched off like that, because it’s decentralised. And this is where the AI/internet analogy stops. Assuming frontier models continue to meaningfully outperform open-source alternatives, access to the frontier remains crucial – but control over this access may be concentrated in the small number of companies building frontier AI and in the country or countries in which those companies are based.

At least in the near term, Australia is unlikely to build its own frontier AI. We don’t have the density of talent or the colossal amounts of capital that appear to be necessary.** Only two countries are running that race: the United States and China. Australia will generally choose American models. But American AI companies exist under American law. They can be controlled and coerced by American governments.

We caught a glimpse of this in March, when the Pentagon declared Anthropic a "supply chain risk". And again in April, when Anthropic gave early access to a model with unprecedented cyber capabilities, Claude Mythos Preview, to selected organisations, all of whom were US-domiciled or headquartered; no Australian business or government agency was on the initial list. And we saw yet another glimpse in June, when the US government imposed export controls that denied foreign users access to Fable 5 and Mythos 5 (though in practice this forced Anthropic to suspend access for all users).

So the question I keep coming back to is: if middle powers are bound to import frontier AI, how do we ensure continued access?

This brings us back to compute. In a loosely analogous way to how Australia gave up on acquiring nuclear weapons yet maintained some influence over nuclear politics through its preponderance of uranium deposits, we may not build our own frontier AI but we can host a substantial portion of the vast amounts of compute needed to train and/or run these models. We could use this as leverage for frontier access – an idea that’s being called ‘compute-for-access’ in AI policy circles. Very, very roughly, this could look like the Australian government striking deals with hyperscalers and major AI labs: we fast-track approvals and facilitate grid access in return for the labs guaranteeing Australians access to frontier AI on equal terms to the US market, if not the US government. Of course, such an agreement doesn’t bind US administrations. But once billions of dollars of CapEx are sunk into the Australian continent, and if we are indeed contractually permitted to throttle data centres’ power supply or even step in and repurpose scarce compute entirely if we lose frontier access, it does incentivise the labs to advocate for our interests in Washington. 

Now, the purpose of this podcast series isn’t to advocate for compute-for-access specifically. It’s just one example of the strategic benefits compute might offer; there are credible alternatives.

And as always, the devil will be in the details.

Which is why I'm doing this as a series of interviews – so I can explore the issue from different perspectives, and so we can see where the guests’ answers line up and where they don’t.

The compute question may be the most important question in Australia at the moment, and I don't think it's being covered with the rigour it deserves.

So for the next month, this is pretty much all I'm doing. I'll release episodes across the month, as I record them, rather than all at once, because things are moving so quickly. First up is Andrew Charlton, who is arguably the key member of the Australian government shaping our AI strategy. More guests will follow.

Subscribe to the podcast wherever you get your podcasts so you don’t miss the episodes as they drop.

See you for the first episode.

*As of the 31st of March 2026, Australia had only three data centres exceeding 100 megawatts. It has no gigawatt-scale data centres in operation, though at least one is being planned.

**In terms of the capital required, one recent estimate put the costs for a coalition of middle powers to develop frontier AI at $500 billion USD over 4 years. For reference, that is about 12 times Australia’s annual defence budget, or roughly double what AUKUS will cost, squashed into a fraction of the time (4 years rather than 30).