Hollywood director James Cameron warns of ‘Terminator-style apocalypse'
In an interview with Rolling Stone, the 70-year-old filmmaker warned that humanity could face extinction if it was used with inappropriate intent.
'I do think there's still a danger of a Terminator-style apocalypse where you put AI together with weapons systems, even up to the level of nuclear weapon systems, nuclear defence counter-strike, all that stuff,' Cameron said.
'I feel like we're at this cusp in human development where you've got the three existential threats: climate and our overall degradation of the natural world, nuclear weapons, and super-intelligence.'
'The theatre of operations is so rapid, the decision windows are so fast, it would take a super-intelligence to be able to process it, and maybe we'll be smart and keep a human in the loop.
'But humans are fallible, and there have been a lot of mistakes made that have put us right on the brink of international incidents that could have led to nuclear war. So I don't know.'
Uncertainty has been a recurring theme in ongoing conversations about the role the technology will play in the future, though experts are certain of one thing: it's only a matter of time until AI is integrated into nuclear weapons.
Last month, at a talk at the University of Chicago for Nobel laureates, Stanford University Professor Scott Sagan said that we were entering a new world of AI that was not only influencing our daily lives, but also 'influencing the nuclear world we live in', as reported by Wired.
Similarly, Bob Latiff, a member of the Bulletin of the Atomic Scientists' Science and Security Board, said he believed AI will be as commonplace as electricity.
'It's going to find its way into everything,' he said.
Where making decisions about nuclear weapons is concerned, experts maintained that ensuring human control remains front of mind is key.
Director of Global Risk at the Federation of American Scientists, Jon Wolfstal, also pointed out one major flaw with integrating artificial intelligence into nuclear weapons: nobody really understands what AI is.
'A number of people have said, 'Well, look, all I want to do is have an interactive computer available for the President so he can figure out what (Vladimir) Putin or Xi (Jinping) will do and I can produce that data set very reliably,' he said, referring to large language models.
'I can get everything that Xi or Putin has ever said and written about anything and have a statistically high probability to reflect what Putin has said.
'I was like, 'That's great. How do you know Putin believes what he's said or written?'
'It's not that the probability is wrong, it's just based on an assumption that can't be tested.'
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
AU Financial Review
12 hours ago
- AU Financial Review
Australian authors go to war against AI
Thomas Keneally says it's life or death stuff, while Richard Flanagan says the Productivity Commission 'doesn't give a shit about Australian culture'. The Booker Prize winners and other prominent Australian authors have slammed suggestions that policymakers should explore ways for tech giants to train AI models on unlicensed creative work. Publishers warn that such copyright exceptions would limit their ability to develop new talent.
AU Financial Review
12 hours ago
- AU Financial Review
Business warns of ‘union veto' of AI use
The country's four major business groups have united to oppose a key government agency's backing for workers to co-design the rollout of artificial intelligence, fearing it paves the way for a union veto over the technology. Jobs and Skills Australia's landmark report on generative AI on Thursday recommended a whole-of-government institution to steward the transition to AI, a national compact to set out principles of AI implementation and support for workers 'co-designing' the technology.
The Age
15 hours ago
- The Age
Friend or foe? The AI bloodletting has begun in Australia
Telstra boss Vicki Brady, though less vocal on the company's AI developments at this week's results, was similarly forthright at its recent strategy day. 'We see lots of potential across those areas … customer engagement, how we operate and manage our network, how we develop software and manage our IT environment, how it supports back of office for us where you tend to have manual processes.' While it sounds like a great opportunity for Australian business, it sounds rather alarming when viewed from the vantage of their employees. After all, both are relatively low-growth businesses investing heavily in AI. Will this investment pay off by boosting worker productivity, or by replacing them? 'CBA publicly preaches productivity and innovation while quietly eroding local jobs. This hypocrisy cannot go unchallenged,' Finance Sector Union national secretary Julia Angrisano said after the bank's record $10 billion profit this week. The Australian Council of Trade Unions has demanded that employers guarantee workers' job security before introducing artificial intelligence to protect against jobs carnage. Local academics used research by the International Labour Organisation to translate its findings on AI job losses to Australia. They came up with a startling forecast of Australia's AI future in 2050: 32 per cent of current jobs in Australia could be done by AI. 'But that doesn't mean 32 per cent of people will lose their jobs overnight,' Victoria University academics Janine Dixon and James Lennox said in a report posted to The Conversation last week. 'It will take time for AI capabilities to be installed, giving people time to train for alternative careers. Much of the impact is likely to be years away.' This timeframe gives AI a lot of time to move beyond relatively low-level tasks, like replacing basic call centre work, to replacing white-collar jobs – like the software developers who make it. Loading So what does Atlassian co-founder Mike Cannon-Brookes think the impact will be of the AI transformation on the company he built alongside Farquhar? It is in a frenzy of AI upgrades of its own products and surely looking at the productivity benefits. Cannon-Brookes sees a bright future despite AI's coding adeptness. It appears that the famously prescient 2011 claim by US billionaire venture capitalist Marc Andreessen that 'software is eating the world' still holds. 'Do I think there will be far less developers in the world five years from now? No, I don't think so,' Cannon-Brookes told investors on the company's earnings conference call last week. 'And yes, we're still hiring lots of engineers and developers with the growth of the business.' Cannon-Brookes' argument is simple: the world will need far more software and AI means it will be cheaper and easier to extend its development beyond corporate tech teams to the actual business itself. 'Whether they're in finance or HR or marketing, there's going to be a lot more people creating software,' he says. Mind you, Atlassian has a lot riding on this version of the future. Its business is literally built on managing the workflows and projects for this sort of development. The Farquhar and Cannon-Brookes fortunes will rapidly dwindle if this development can be done by an AI bot instead of teams of employees. But even AI's transformation of low-level customer work – like call centres – is not necessarily seen as a bad thing for local jobs. It could represent a boon for our country, says KPMG's chief digital officer John Munnelly. 'A lot of the stuff that AI is improving is the tasks we used to offshore, like call centre work,' he says. 'There's a really great opportunity for the Australian economy with AI' KPMG's chief digital officer John Munnelly The productivity dividend that Farquhar mentioned could actually make a lot more of this work viable here. 'There's a really great opportunity for the Australian economy with AI,' Munnelly says. Loading But the interesting stuff is already happening further up the wage chain – like KPMG's new AI tax tool that allows its executives to vastly accelerate the delivery of first-draft advice to clients. 'What used to take us two weeks to go and prepare – if a client's in the middle of a deal – now we can literally get it out the door in a day,' Munnelly says. It was left to KPMG chief executive Andrew Yates to address the conundrum this poses. What will this KPMG employee do with the nine days that would have previously been spent on this work? 'I think our current hypothesis is that what we do will change. But AI and the technology we've got will generate so much more data that our work will change from collating that data to really assessing, analysing, presenting, interpreting much more data than is currently available,' he says. 'There will be a real need for that insight and technical understanding of all the data that's produced.' As for the analysts trying to make sense of the AI talk which is starting to creep into earnings season speeches and rising costs, there is a more prosaic question. 'Companies have been keen to point out their investments in AI, but when will we see it translate to the bottom line?' asked UBS strategist Richard Schellback. Even Comyn, who packed more than a dozen AI references into his introduction to the bank's full-year results, came up with a cautious answer. He foresees a more effective workforce, producing higher quality work with both revenue and cost out opportunities. But he does not expect this to come easily. 'You can imagine that there are some much more efficient ways of delivering some of the things we currently do. But I do think that's going to take some time, like some years to work through some of the accuracy and quality that's required.'
