Mark Zuckerberg positively beamed as he gazed into the future. Relaxed and upbeat, the T-shirt clad Facebook founder looked into the camera and announced his ‘vision to build Personal Superintelligence for everyone.’
Hurtling towards us, he vowed in the Instagram video posted last month, was ‘a new era of personal empowerment.’ Limitations on science and health would be shrugged off and mankind would be, ‘spending more time on creativity, culture, relationships, and enjoying life.’
So far, so positive. And so keen is Zuckerberg to realize his vision, he is luring top Artificial Intelligence researchers with pay packages of up to $300 million over four years.
But, save for imagining interactive intelligent glasses that will ‘see what we see, hear what we hear,’ and become our ‘primary computing devices,’ the 41-year-old was notably short on details.
And, according to a host of world experts who have spoken to the Daily Mail, that detail is where the devil lurks.
Because where Zuckerberg’s future is bathed in a halcyon glow, any light they see is the flicker of the flames in which, they predict, civilization will be engulfed if Personal Superintelligence is unleashed upon it.
Put bluntly, one said, introducing Superintelligence as an aid to our daily life is like, ‘inventing nuclear bombs because they could help heat your house more efficiently.’
As far as some of the greatest minds in the field are concerned, the potential for this unregulated, unpredictable and entirely uncontainable power is simply too great to ignore.
Zuckerberg (pictured trying on Orion augmented reality glasses in 2024) is luring top Artificial Intelligence researchers with pay packages of up to $300 million over four years
Put bluntly, one expert said, introducing Superintelligence as an aid to our daily life is like, ‘inventing nuclear bombs because they could help heat your house more efficiently’
And Zuckerberg’s notion that we can somehow harness it in a master/servant-style dynamic is, they said, delusional.
One expert told the Daily Mail it was like giving a monkey a human as their own ‘personal superintelligence’ – failing to note that the human could lock the monkey in a cage, starve it or just behead it.
Nick Bostrom, the founder and director of the Future of Humanity Institute at Oxford University, believes that Zuckerberg is significantly downplaying the seriousness of Superintelligence – an as-yet unachieved milestone.
Bostrom, whose groundbreaking 2014 book, ‘Superintelligence,’ was described by Open AI founder Sam Altman as ‘the best thing I’ve seen on this topic,’ has described the technology as ‘an intellect that is much smarter than the best human brains in practically every field, including scientific creativity, general wisdom and social skills.’
Computers will be able to think for themselves, work together and, with the aid of robots, perform currently unimaginable tasks. And while AI is a tool for use by humans, Superintelligence will act alone.
Speaking from his home in the British countryside, Bostrom told the Daily Mail, ‘If it actually is developed, it will be more radically transformative than just another cool consumer appliance.
‘It’s a much more fundamental transformation of the human condition.’
At best, Bostrom said, Personal Superintelligence would allow people access to their own army of assistants and advisors, which would mediate and lobby on their behalf.
At its more extreme, no one would work, he said, suggesting our existence could be like that of ‘old British aristocrats… hunting, having a drink at night with their friends, or pottering around in their gardens.’
Hurtling towards us, he vowed in the Instagram video posted last month, was ‘a new era of personal empowerment’
But, save for imagining interactive intelligent glasses that will ‘see what we see, hear what we hear,’ and become our ‘primary computing devices,’ Zuckerberg was short on details. (Pictured: Ray-Ban Meta Smart Glasses)
Bostrom has described the technology as ‘an intellect that is much smarter than the best human brains in practically every field.’ Computers will be able to think for themselves, work together and, with the aid of robots, perform currently unimaginable tasks
If that utopian outcome seems improbable, that’s because, according to many, it is.
The majority of the experts who spoke to the Daily Mail were clear: for all Zuckerberg’s breezy promises, the most likely outcome if he succeeds is nothing short of humanity’s destruction.
David Krueger, an assistant professor of AI policy and safety at the University of Montreal, said: ‘What Zuckerberg is talking about is not coherent, or he’s just using Superintelligence as a branding term to describe something completely different from what that word actually means,’ he said.
The reality, he said, is that the technology being built, ‘poses an existential risk… is going to completely reshape the world… is going to at a minimum take everyone’s jobs and end up running society and lead to an unprecedented concentration of power.
‘It is probably also going to literally lead to the end of humanity.’
Nate Soares, whose book, ‘If Anyone Builds It, Everyone Dies,’ is out next month, told the Daily Mail he has been warning tech titans – including Sam Altman and Elon Musk – about the risks for years.
He has, he said, gone unheeded: ‘Before ChatGPT was a glimmer in Sam Altman’s eye – even before OpenAI was a glimmer in Sam Altman’s eye – I talked to Sam and Elon and gave some suggestions, that they did not take.’
Soares said it was entirely possible a Superintelligence reasons that ‘a human takes 100 watts of power to run, and there’s more efficient things it can do with those 100 watts.’
Superintelligence could work to improve solar power, for instance, and harness all the sunlight, leaving mankind to wither and die in darkness. It could also drain the seas.
The majority who spoke to the Daily Mail were clear: for all Zuckerberg’s breezy promises, the most likely outcome if he succeeds is nothing short of humanity’s destruction
Where Zuckerberg’s future is bathed in a halcyon glow, any light experts see is the flicker of the flames in which, they predict, civilization will be engulfed if Personal Superintelligence is unleashed upon it
Soares explained: ‘If you are trying to do computation on the surface of the earth, your limit is not energy generation; your limiting factor is actually probably cooling.
‘One guess is it figures out fusion, and then it starts boiling off the oceans to cool the fusion reactors so it can do even more fusion and run even more computation while it’s trying to figure out how to most efficiently rip apart some planets and wrap them around the sun so it can collect a whole solar unit of energy.
‘It may sound sci-fi, but the stuff that humans are doing relative to chimpanzees sounds sci-fi too.’
Soares continued: ‘AI just does its own thing, its own weird thing that we didn’t understand – and it doesn’t care about you.’
Katja Grace, co-founder of the research organization AI Impacts, whose thinktank regularly surveys the world’s top AI minds, agreed: ‘The difficulty is that we’re taking a random network and then nudging it an unimaginable number of times until it does what we want and the result is unpredictable.
‘We’re growing them rather than building them.’
Indeed, when it comes to regulating this technology’s development, Krueger said there, ‘needs to be an international agreement, and it needs to happen like yesterday. This should be the number one foreign policy priority of every nation.’
And there is no time to waste.
Because, according to Soares, what Zuckerberg is unwittingly doing is nursing a lethal predator. He said: ‘We essentially have baby raptors; they’re still cute. And they’re going for the door latch.’
