The AI revolution came later than expected but also had a more immediate impact than I had imagined.
I remember the moment I realized how transformative AI was going to be. It was the mid-1990s and I was about to graduate with my undergraduate degree in computer science.
When deciding to pursue my law or computer science degrees, I saw a proposal for a research project identifying pairs of human chromosomes using computer vision – a type of AI that allows computers to interpret and understand information about the world through images.
The fact that it was possible to automate this complex task made me realize at the time that this technology was tackling one of the fundamental problems in science and that, when it worked, it would have an impact huge impact on the way we live.
I underestimated both the time it would take for the technology to mature and the impact it would have once it did.
Of course, AI didn’t come out of nowhere, it relies on much of the technology that humans have developed before: the Internet, ICT, electricity and even the steam engine.
But AI’s flexibility, generality, and ability to do things that would traditionally require a lot of human labor make it truly revolutionary.
Over the years, my research team at the University of Adelaide has built a reputation for its specialist capabilities in computer vision. It started with a research center with four members, and grew into one of the best groups in the world with over 130 members when I left as director.
At first, it was difficult to convince people that AI was going to be transformative. Grants and partnerships with a few international companies who understood the value of AI were essential to our growth, and I was ultimately able to transform this center into an institute in 2018.
Today, the Australian Institute for Machine Learning (AIML) is home to approximately 180 researchers, students and engineers.
Our collaborators publish new research at the cutting edge of machine learning science and work with local and multinational companies to create innovative AI solutions for all kinds of problems.
The institute has enjoyed strong financial and policy support from the South Australian Government, but without our founding partnership with Lockheed Martin, we may never have been able to get it off the ground.
Computer vision is an area of AI that is increasingly integrated into overall defense and strategic intelligence capabilities, and Lockheed Martin has seen the value of investing in fundamental research.
They have invested in developing an AI capability. They wanted to engage and have access to qualified staff, a talent pool of smart students, new ideas and the latest developments, and have their collaborators in our building so they could work with the best group of people. ‘Australia.
That an American defense company enabled the creation of the Australian Institute for Machine Learning is not lost on me. But no Australian company was sophisticated enough to engage in this type of industry-academic research and development model.
Russia’s invasion of Ukraine gave us a glimpse of what future war will look like. With significant help from its Western allies, Ukraine has created its own AI geospatial intelligence platform to fend off a much larger adversary.
This includes computer vision to analyze vast satellite, aerial and ground imagery to understand troop and vehicle movements, and natural language processing to autonomously mine large amounts of unencrypted radio communications for intelligence. exploitable.
As Russia is learning, no amount of firepower or force on the ground can beat data-driven agility.
Just over two years ago, the Australian government announced a $360 billion trilateral AUKUS security deal that could see the first nuclear-powered submarine arrive in Australian waters by end of the next decade.
But we urgently need to increase our national capabilities in AI and autonomy. China is miles ahead of us in AI capability and is deploying it flexibly across its entire military system.
In April, the Australian government published its Strategic Defense Review; The 112-page public report mentions artificial intelligence only once, when it notes that AUKUS Pillar II “will develop and deliver advanced capabilities in areas such as artificial intelligence, hypersonics and of the maritime domain”.
AI is not something we can buy off the shelf from our international allies and expect it to work once plugged in. Technology that can be adopted elsewhere is derivative and offers no competitive advantage.
The world is spending heavily on AI, and our allies and adversaries are making strategic decisions to invest in their core national AI capabilities by investing in their smartest people.
The US and China are obviously the world leaders, but the UK, France, Germany and South Korea have all invested billions of dollars, if not several billion dollars.
Singapore, with a population only slightly larger than that of Sydney, has a national AI strategy supported by approximately A$740 million in fundamental research, translational research and industry-research collaboration .
How can we address our national AI deficit? In many ways, it’s simple: we need to start funding it properly.
But Australia also needs a strategy to engage with the best minds in AI so it can build a world-leading AI defense capability. One way to do this is to conduct defense AI research in an unclassified environment.
Defense can’t just recruit the best AI talent, because the best people are at our leading universities, and they won’t go to work on the wrong side of a 10-foot classified fence under a publications embargo .
Scientists need an academic environment of collaboration and open competition among peers to remain at the global forefront in their field. This is essential for their career and performance.
Defense technology leaders must understand that great technological innovations are developed in the open, not in a classified laboratory, disconnected from the rest of the world. A top-secret research environment often results in second-rate results. Entropy increases in a closed system.
So how can we conduct cutting-edge AI research in a way that meets Australia’s defense needs?
It goes something like this: Someone in defense proposes their toughest technology priorities in a secret, fully classified area. They then create an unclassified euphemism for this problem.
University researchers set out to solve this euphemism using the world’s best AI technology. And when researchers have solved the problem, defense can then bring the technology back beyond the 10-foot barrier and into the classified laboratory.
This idea is not radical and it is not new. The Defense Advanced Research Projects Agency (DARPA) has been conducting high-risk, high-reward research in the United States for nearly 65 years and is partly responsible for some of the biggest landmark inventions of the last half century: the Internet, computers personal, GPS. , stealth technology and even Covid-19 vaccines.
DARPA understands the value of basic research and “fully supports free scientific exchange and dissemination of research results wherever possible.”
A member of my team recently asked me what worries me most about Australia’s lag in AI and what the next five to ten years might bring for defense technology.
This is an important question that is almost impossible to answer. To borrow a NASA expression, the future impact of AI is one of the great unknowns. But I know that sovereign AI capability is absolutely critical to Australia’s future.
Professor Anton van den Hengel is an AI researcher at the University of Adelaide and founding director of the Australian Institute for Machine Learning (AIML). Currently, he is Director of the AIML Center for Augmented Reasoning and Director of Amazon’s Machine Learning Research Team at Adelaide’s Lot Fourteen Innovation District.
Do you know more? Contact James Riley via E-mail.