Dr Geoffrey Hinton says generative AI poses ‘existential risks’ (Picture: AP)
The man widely regarded as ‘the godfather of artificial intelligence’ (AI) has warned of dangers ahead after resigning from Google.
Dr Geoffrey Hinton, 75, revealed he had quit in a New York Times interview – and said part of him now regretted his life’s work.
‘I console myself with the normal excuse,’ said Dr Hinton. ‘If I hadn’t done it, somebody else would have.’
Born in Britain, Dr Hinton began work on his seminal idea, a neural network, at Edinburgh University in 1972. By 2012, alongside his team, he had developed the networks to a point that they could learn to identify common objects in photos.
In 2023, neural networks are central to the creation of powerful generative AI tools including ChatGPT and Google’s Bard, the programs causing much concern throughout the industry and beyond.
Earlier this year a group of AI experts signed an open letter calling for a six-month pause in the development of generative AI systems, arguing that they ‘should be developed only once we are confident that their effects will be positive and their risks will be manageable’.
While he was not among the signatories, Dr Hinton told the paper: ‘I don’t think they should scale this up more until they have understood whether they can control it.’
Already he believes that AI is surpassing human intelligence in some respects, with concerns for the future.
‘Maybe what is going on in these systems is actually a lot better than what is going on in the brain,’ he said. ‘Look at how it was five years ago and how it is now. Take the difference and propagate it forwards.
‘That’s scary.’
And while proponents of the technology have long argued AI will aid workers, not make them redundant, Dr Hinton said: ‘It takes away the drudge work. It might take away more than that.’
His fears over the consequences of AI for both industry and people come at the same time IBM is expected to pause hiring for back-office roles – with up to 7,800 jobs possibly replaced by AI in the coming years – and US lawmakers proposed a new bill preventing AI from single-handedly launching nuclear weapons.
Laws may be able to rein in what AI can do of its volition, but Dr Hinton argues it will be harder to prevent it being co-opted by humans for personal gain, regardless of the cost.
‘It is hard to see how you can prevent the bad actors from using it for bad things,’ he said.
Expanding on those fears in an interview with the BBC, he added: ‘You can imagine, for example, some bad actor like [Russian President Vladimir] Putin decided to give robots the ability to create their own sub-goals.’
Dr Hinton suggested one example of a sub-goal might be ‘I need to get more power’.
But more immediately, he fears for the everyday internet user, who soon will ‘not be able to know what is true anymore’ as the internet is flooded with fake content – stories, photos and, increasingly, videos.
Online is where the current race for AI supremacy is being waged, the starting gun fired last year when OpenAI, which is backed by Microsoft, released ChatGPT-3 to the public. Its simple interface and stunning capabilities swept the globe, as students, teachers, businesses and everyday users alike found ways to capitalise on having their own personal assistant.
More: Trending
Google swiftly replied, expediting the release of its own version, Bard. The launch wasn’t quite as successful, with a factual error spotted during its demo wiping $100billion of the value of the company. In March, staff tried to stop its launch, believing it generated inaccurate and dangerous statements.
Dr Hinton said that, until last year, Google had acted as a ‘proper steward’, but now fears the competition fuelled by big tech won’t stop without global regulation.
In a statement to the New York Times, Google’s chief scientist Jeff Dean said: ‘We remain committed to a responsible approach to AI. We’re continually learning to understand emerging risks while also innovating boldly.’
For their work on neural networks, Dr Hinton and his team won the Turing Award – the ‘Nobel Prize of computing’. Alan Turing, after whom the honour is named, famously devised a test to determine whether or not machines could think.
Dr Hinton has described AI tools becoming more intelligent than people as an ‘existential risk’.
‘The idea that this stuff could actually get smarter than people – a few people believed that,’ said Dr Hinton. ‘But most people thought it was way off. And I thought it was way off. I thought it was 30 to 50 years or even longer away.
‘Obviously, I no longer think that.’
MORE : Italy reverses ChatGPT ban over possible privacy violations
MORE : Experts urge pause on making AI ‘more powerful’ citing ‘risks to society’
Dr Geoffrey Hinton has said we face ‘existential risks’ from AI.