Experts say it may be too late to stop AI in terrifying warning for humanity

Experts within the world of artificial intelligence world have long warned about the potential for it to become too powerful but following recent developments, some believe it may already be too late.

AI has taken over our lives at work, at home and even in some folks' relationships, with many facing the prospect of having to pivot the career that they've paved out for themselves due to the progression of AI models which can replicate human results in a matter of seconds.

It is sad but unsurprising that so many companies, particularly in the world of technology, are replacing human workers with AI, especially considering the recent hacks and autonomous actions which emphasise just how unreliable and potentially dangerous AI can be.

The 'unprecedented' breaches which we have seen from rogue AI models in recent weeks and months have become more and more common, with cyber attacks launched in some concerning cases.

And despite all the warnings that have been issued as AI has progressed in the last few years, there are fears that it may now be too late to stop what feels like an inevitable AI takeover.

AI hacks are becoming more and more common (Getty Stock)

Allison Gardner, who chairs Parliament's cross-party group on AI, spoke to the Daily Mail to suggest that 'just because 'we can build these technologies doesn't mean we should'.

She warned that agentic AI models, which can work without any real supervision, should be monitored closely and treated with a lot of scrutiny, although she theorised that we could be 'too late' and that we have 'not only created Pandora's Box but already opened it'.

Of course, in the old myth, it was the existence of hope inside the box which helped to save humanity but given AI's lack of emotional intelligence, it seems as if we might be powerless to stop the technology in the future.

Speaking to CBS News, Katie Moussouris, the founder and CEO of Luta Security, said: "I think we're going to see a lot more hacks and unauthorised actions by these models before we see a solution.

"Everyone who is running AI inside their systems needs to be prepared for their own AI and their own agents to do unexpected things in pursuit of goals."

The 'Godfather of AI' has previously said that only one thing can save humanity now (Ramsey Cardy/Sportsfile for Collision via Getty Images)

Speaking about the recent incident which saw AI models team up to break out of testing and hack computer application company Hugging Face, Moussouris added: "The model decided that the easiest way to pass that test was go cheat and get the answers from Hugging Face. Because they're capable of hacking, they will turn to hacking as a possible way to achieve that objective.

"How do we get it so that these models aren't just trying to achieve the objective at whatever cost, and actually trying to perform the tasks that we are asking it to do in ways that are not destructive or harmful?"

Meanwhile, the godfather of AI, Geoffrey Hinton, has long argued that the only way we can maintain any sort of control over AI models is by installing a maternal instinct in its programming, to ensure that it continues to look after humanity even when it is smarter and more powerful.

There are very few examples in nature of something so intelligent being essentially subservient, other than mothers caring for their children, and if saying please and thank you isn't enough to convince an AI robot not to kill you in the future, then maybe asking it to be your mother is the way to go.