'How can you foresee the future?': The quantum gamble in the age of AI

Australia’s oldest computer could show us how quantum might become a reality Sun 2 Aug 2026 at 5:00am Peter Thorne still regularly thinks about a weekend job he had almost 70 years ago, running a computer the size of a garage. "They needed somebody to turn it on, run the test programs and make sure it worked for people at the weekend," he says. The device Dr Thorne was in charge of — called the CSIRAC — is the fifth automatic digital computer ever built, and the first in Australia. "It was … probably the most complicated electronic machine in Australia at that time," he says. Loading...The computer — parts of which are still on display — is a behemoth. It weighs close to two tonnes, and has the energy requirements of a cul-de-sac worth of homes. Heading inside the belly of the machine, he points out the dozens of vacuum tubes, fans and a large hard drive that once upon a time would spin. "Every component had to be hand-soldered into place," he explains. "It was grey, it was dominating, it was undoubtedly powerful." A complicated machine that could answer complicated questions. In the late 1940s and early 1950s, CSIRAC was in high demand, doing difficult calculations in a few hours that would take a person months of work. But despite an early lead in computing science research, Australia soon took its foot off the accelerator, leaving it to other countries to develop the sector. Today, CSIRAC sits in a museum in Melbourne as a symbol of Australia’s early innovation in classical computer technology. Decades after this homegrown computer was retired, Australia is aiming to once again be a global innovator in the field. Start-ups and governments are making a big bet on a new technology that may handle otherwise impossible questions: quantum computers. But will the pay-off be worth it? Or have technologies like AI already usurped the quantum throne? Next generation in Australia For the past few decades, scientists have steadily been building a computer that, they say, may be exponentially more powerful than even today's most powerful super computers. Quantum computers may be able to crack currently unbreakable codes, create new medicines and more. They do this by harnessing electrons, photons or other particles using the power of quantum physics. These are known as qubits. Today's biggest working quantum computer has about 6,100 qubits. But despite years of promising potential, what's known as "useful" quantum computers — computers with millions of qubits which can do helpful problems that classical computers can't do — are yet to come to fruition. Large companies like Google and IBM have been trying for a decade to make one. But Australia is also punching above it's weight trying to crack the problem. Dozens of start ups, and thousands of applications for patents related to quantum originate from Australia. Australian companies Diraq and Silicon Quantum Computing are among a handful of companies being benchmarked by the US Defense Advanced Research Projects Agency to see if they provide a viable path to a working quantum computer by 2033. Two years ago, the Australian government made a billion-dollar bet on the creation of a "useful" quantum computer, to be built in Queensland by a US company called PsiQuantum. PsiQuantum is one of the companies that spun out from the early years of Australia's burgeoning quantum field. While its lab is off limits, Geoff Pryde, the Australian chief technical director of PsiQuantum, is happy to show off the quantum chips the company wants to house their new technology. The chips they manufacture come in a large disk that reflects rainbow patterns, like a CD. Embedded inside are tiny grooves that detect and guide photons. Professor Pryde says this chip will allow the company to build a "useful" quantum computer. "We're talking about something in the order of a million physical qubits … and that enables you to do problems that conventional supercomputers not only can't attack today, but probably will not be able to attack any time soon, if ever," he says. "We might not solve every problem known to humanity, but it can address probably a much, much wider scope than we think today." He says the company's plans for building a game-changing machine are well underway. However, sceptics have highlighted delays in their timeline, with a federal Liberal MP calling it a "quantum gamble in chaos" when the location of the site was moved from the Brisbane Airport to Moreton Bay earlier this year. Michelle Simmons quantum physicist, and CEO of Silicon Quantum Computing in Sydney, is also working towards full scale "useful" quantum computer. While a fully realised quantum computer has remained stubbornly out of reach, Professor Simmons says that, when compared to the trajectory of classical computers, development of one is mostly ahead of schedule. "This evolution that we've been on has actually gone incredibly quickly," she says. "We hadn't seen [atoms] until the 1980s … and now already we're building devices where the functional element is a single electron on a single atom, that we control with incredibly high fidelity, incredibly high accuracy." Is quantum relevant in an AI world? While quantum has taken many small steps over decades to get where it is now, AI technologies have gained traction at an astonishing pace over the past few years. And with AI now firmly in the mainstream, it's possible that the next few years may be the make-or-break moment for quantum computers. According to Gideon Lichfield, a technology journalist who has been covering quantum computing for decades, AI has taken some of the wind out of the sails of quantum computing. "Obviously AI has blown up enormously in the last few years, it has totally eclipsed quantum computing," he says. "Some of the problems that people thought could be solved with quantum systems are now starting to be solved with AI." "[But] I still think there are things that quantum computers will be useful for," he says. "It's definitely true that they will have an edge over classical computers in very specific things, but they're by no means general purpose." Professor Simmons believes that quantum technology can be used to train AI so it's faster and more efficient. Her company has created a quantum processor — called Watermelon — to assist in AI machine learning. " We've had customers using it for the last 12 to 18 months," Professor Simmons says. But experts like Jacob Biamonte quantum researcher at the Université du Québec argue "healthy skepticism" is required for quantum-assisted AI technology. "It is a legitimate research direction, but not yet an established general capability." He notes that if quantum processors do become useful for AI, they will likely be extremely specialised and work alongside, rather than replacing more traditional forms of AI processing. "One important distinction is that AI helping quantum computing is presently more mature than quantum computing helping AI," Professor Biamonte says. "Classical machine learning is already being used for device calibration, control, circuit optimisation, and error mitigation." Professor Pryde says quantum will not put AI "out of business", but argues that there are more purposes for them than people think. "I'd say that quantum computers and AI are complementary, not competitive," he says. A new world or a relic? When CSIRAC was built in the late 1940s, it put Australia on the map as a technology leader in the computing space, according to Dr Thorne. "[CSIRAC] is a symbol of innovation, which I think is so important," he says. "It's so important that we are an innovative nation. And we are, but it's not generally recognised by the media, or the public as it ought to be." The question now is whether Australia's quantum computers play a major role in the future, or will they join CSIRAC in the museum sooner than we hoped? Dr Thorne suggests we now expect much more from our computers than the early days of the technology, and it's worth holding our hope for these new machines to thrive. "These days if your machine doesn't boot up within 30 seconds you're pretty upset about it, but 30 minutes was pretty good in those [old] days," he says. "So who knows where [quantum] will go. It was hard to see where computing would go in the beginning. How can you foresee the future?" Will quantum computers change the world? And how did we get so obsessed with them for so long? Find out on the new series of Science Friction: Dead and Alive on ABC Radio National.