Australian innovators should stop trying to create sovereign artificial intelligence models, a US expert has warned, and instead use the tools made in other countries to boost productivity.
Subscribe now for unlimited access.
or signup to continue reading
See subscription optionsAuthor and academic Jeffrey Ding issued the warning at the KWM Digital Future Summit on Wednesday, saying Australian researchers didn't need to create the "next great breakthrough" to succeed in the technological race.
The warning comes after the Productivity Commission recommended against introducing mandatory guardrails for AI to extract the most from it, and federal Treasurer Jim Chalmers said Australia would plot a "middle course" to balance the technology's benefits and risks.
Generative AI technology has emerged as a major productivity tool since the launch of OpenAI's ChatGPT in 2022, inspiring competition among large tech firms and countries alike.
But while China and the US continued to treat the battle like a race, Australian companies could extract the greatest benefit from AI by deploying it throughout the economy, Mr Ding said.
"In the long run ... for countries like Australia and middle powers, it's not worth it to chase sovereignty because it's really hard to attain.
"Sometimes we start with the wrong function where it's about seeing can Australia come up with the next great breakthrough like wi-fi.
"The real function is can Australia be ahead of other countries in adopting the next wi-fi at scale to drive productivity growth?"
The George Washington University assistant professor said AI looked like it could be a force for major change in a way similar to the steam engine, electricity and the computer had been in the past.
But the technology would need to be deployed throughout industries, outside the information technology world, to make the greatest impact on economies and productivity, he said.
Preventing individual states from introducing AI restrictions and introducing a national AI regulator could also help more companies experiment with AI software, Mr Ding said.
"In the long-term, having a government organisation that can help with AI safety benchmarks and testbeds and creating trust around the technology ... would be useful," he said.
The federal government issued voluntary rules for the use of AI in September and held a public consultation into mandatory AI guardrails and laws.
The Productivity Commission warned against introducing limits on high-risk AI uses, however, warning that poorly designed regulations could limit innovation.
Australian Associated Press
