On AI ‘we can’t compete on scale, only expertise’

Dr Jon Whittle

A framed intergalactic trader, a prison planet and the tantalising prospect of escape is what started it for me. I first came across Artificial Intelligence (AI) as a teenage game designer, filling the hours between school and dinner using early natural language processing (NLP) to build the world of ‘Setup’.

It wasn’t a bad way to spend time as a 14-year-old, and it even made me a bit of money.

‘Setup’ was an early text adventure computer game. It used NLP, an element of AI, and was like nothing I had used before.

I could create magical virtual worlds where game players could navigate simply by typing commands in everyday language. PICK UP AXE AND KILL ORC. And the computer understood. The relative simplicity was game-changing.

As a teenager in the 1980s, to interact with a computer so directly felt revolutionary. Fast forward 40 years and that same type of text-to-command interaction is so ubiquitous we barely notice it.

Data61 chief Dr Jon Whittle

After a number of false starts and two so-called ‘winters’, AI has found its way to the mainstream.

NLP is powering the search engines that make navigating the internet possible. Computer vision has given rise to driverless cars and machine learning is changing just about everything we do; from the way we grow food to the way we meet our partners.

According to estimates, the AI sector will be worth $315 billion to the Australian economy by 2028 and $22 trillion to the global economy by 2030.

The Stanford University Artificial Intelligence Index reported that despite the pandemic, there was a 9.3 per cent increase in private investment in AI in 2020, a higher percentage increase than the year before.

Government interest in AI, notably in China and the United States, is also growing – the number of AI mentions in the US congressional record has more than tripled in recent years.

Meanwhile, across the Atlantic, the European Commission announced a coordinated approach to AI policy and investment aimed at making the EU a global hub, and the UK has announced a ten-year plan to become a global AI superpower.

In this context, Australia needs to claim its space carefully because we can’t compete on scale, only expertise.

There is huge potential for Australia to lead globally in areas such as AI for ‘the great outdoors’, leveraging our depth of experience around AI in farming, mining, fishing and forestry, environmental monitoring, and search and rescue.

At CSIRO, we are already using AI to predict bushfires, reduce plastic pollution, protect the Great Barrier Reef, and manage biodiversity.

Australia is well placed to lead in emerging fields like Quantum AI where our two decades of sustained investment in quantum technologies positions us well and has the potential to create $4 billion in economic growth and 16,000 jobs across a range of sectors.

Delivering ethical AI is another clear leadership opportunity. The success of AI – and the ability of this technology to deliver for communities – hinges on the ethics frameworks that underpin their development.

We’ve made a good start in recent years with CSIRO establishing the first AI roadmap for Australia, and the Australian Government putting in place an AI Ethics Framework.

But the next step is to put these into action and help Australian business access the power of AI to drive new opportunities for the nation.

To achieve any of this we must overcome our greatest AI challenge – fragmentation. People talk about talent leaving for overseas, but by far the bigger issue is that the talent we have is largely working in silos and there is no front door for businesses.

This is particularly for small to medium-sized businesses wanting to adopt domestically developed AI solutions.

To tackle this fragmentation challenge, under the AI Action Plan announced in June, the federal government has tasked CSIRO with establishing a National AI Centre for Australia.

The need for a centre like this is widely agreed. Coordination will be key to Australia becoming a leading digital economy by 2030 and for our ability to meaningfully contribute in a competitive AI environment.

According to the Global AI Vibrancy Tool, Australia currently ranks fifth in the world for academic research in AI, but only eleventh for AI innovation.

To create a genuine sovereign capability, we need to lift that performance. And that is a job we can only do together.

At a national level, coming together means we can showcase the depth and seriousness of our expertise and provide a logical focus point for attracting international partnerships and investment in Australian AI research and services.

AI has come a long way in the past 40 years, and it will go even further in the next 40.

For Australia, we need to work together and develop a vision for the technologies and industries we want today’s 14-year-olds to inherit.

Jon Whittle is Director at CSIRO’s Data61 business unit. He is a world-renowned expert in software engineering and human-computer interaction, with a particular interest in IT for social good, AI research in health sciences, and sustainable development.

Do you know more? Contact James Riley via Email.

Leave a Comment

Related stories