Save articles for later
Add articles to your saved list and come back to them any time.
London: Australia has vowed to work with the international community to police the development of artificial intelligence, as Science Minister Ed Husic played down the likelihood of scenarios resembling catastrophic scenes from Hollywood film The Terminator emerging from the technology.
Husic and Deputy Prime Minister Richard Marles announced on Friday that Australia will be a signatory to the Bletchley Declaration after an AI summit held in England’s Bletchley Park, where Alan Turing cracked the Enigma code during World War II.
Minister for Industry and Science Ed Husic.Credit: Dan Peled
Among the 27 countries that signed the declaration were the US, UK and China, after the summit heard from representatives of leading AI companies, civil society groups and academics to consider the risks posed by the technology.
Speaking to this masthead ahead of the summit, Husic said that while there was concern generative AI could go beyond what it is programmed to do, a scenario where humanity loses control of robots was unlikely.
“People within the tech space say ‘yeah that’s a possibility, but exceedingly small’,” he said.
“There’ll be the doomsayers, there’ll be the people who argue that The Terminator and Skynet is around the corner,” he said, referring to the 1984 science fiction film. “But if you look at the way the technology’s developing, those kinds of scenarios aren’t panning out.”
Arnold Schwarzenegger starred as The Terminator, a cybernetic assassin.
Husic said losing control of AI could be mitigated by getting the technology’s regulatory framework correct early on.
Gilbert and Tobin law firm’s digital technology partner, Simon Burns, said that Australia signing up to the agreement was a positive step, but that it was unlikely to make a difference to the domestic regulation of AI.
“What it is doing is making sure we have a seat at the table in terms of the international discussions, which I think is critical because AI is not something that’s just within a silo in Australia, and we need to take a co-ordinated global approach,” he said.
Burns said that rather than implement regulations to stop people from misusing AI, governments needed to enable people to use the technology correctly and understand the risks.
“For Australia, it’s critically important to get that balance right and that we don’t create a regulatory environment which is too rigid and can’t cope with the evolution of tech.”
Husic said one of the more likely worse-case scenarios of AI was the way it could be used by governments and businesses to deprive people of the ability to challenge a decision or have human input.
Asked if he was concerned China could use AI in a hostile way that Western governments would oppose, Husic said it was important to “be prepared and have a robust response to that”.
Husic said people used technology because they believed the government had already considered its application in the same way people consume foods and medicines.
“That’s why I think things like the summit are important to demonstrate that we are thinking it through,” he said.
Katherine Mansted is executive director of cyber intelligence at CyberCX, an Australian cybersecurity firm. She said the Australian government was taking a more proactive approach to AI than it had with the internet, and that the challenge facing regulators was predicting how the technology would impact society.
“We need to bake security and privacy and good governance into AI before it expands and grows further. It’s something that we didn’t do with internet security and cybersecurity, which really has been something of an afterthought,” she said.
“It is hard to predict how developments in AI will impact society, security and the economy if you get a group of AI experts in a room, you will have different predictions from different experts about degrees of threat.”
US President Joe Biden issued an executive order this week which compels companies developing models that potentially pose a risk to national, economic or health security to share training methods with the government.
The US Department of Commerce is drawing up guidelines that will lead to the compulsory watermarking of AI-generated content.
Get a note directly from our foreign correspondents on what’s making headlines around the world. Sign up for the weekly What in the World newsletter here.
Most Viewed in World
From our partners
Source: Read Full Article
