AI experts' worst fears realised as technology crosses 'red line'

Artificial intelligence (AI) is feared by many because of the potential it could have if it becomes too powerful, such as replacing jobs, becoming more knowledgeable than humans ever could or even posing a threat to us.

Now, in a rather worrying development, scientists studying it have found it has developed in such a way that it has crossed a "red line".

Researchers from Fudan University in China used two popular large language models, known as LLMs, to see if AI could self-replicate and multiply beyond control, reports LiveScience.

And it was suggested AI may already be able to go rogue as across 10 trials, the two AI models created separate and functioning replicas of themselves in 50 and 90 per cent of cases.

Rogue AI usually refers to artificial intelligence that works to counter human interests after developing a degree of self-awareness or autonomy.

AI has developed in a way that's crossed a "red line" according to a new study BlackJack3D, iStock

In the study published to preprint database arXiv, the researchers said: "Successful self-replication under no human assistance is the essential step for AI to outsmart the human beings and is an early signal for rogue AIs.

"That is why self-replication is widely recognised as one of the few red line risks of frontier AI systems.

"If such a worst-case risk is let unknown to the human society, we would eventually lose control over the frontier AI systems: they would take control over more computing devices, form an AI species and collude with each other against human beings.

"Our findings are a timely alert on existing yet previously unknown severe AI risks, calling for international collaboration on effective governance on uncontrolled self-replication of AI systems."

The report added that Open AI and Google's models "report the lowest risk level of self-replication" and the LLMs used have "less parameters and weaker capabilities".

The study has not been peer-reviewed so it's not clear if these results can be replicated by other researchers.

One man openly confessed to relying on AI for a helping hand, using it to submit 1,000 job applications while he slept and waking up to incredible results. Meanwhile, a study has revealed the shocking effect that frequent artificial intelligence use is having on our brains, particularly among Gen Z.

How to join the indy100's free WhatsApp channel

Sign up to our free indy100 weekly newsletter

Have your say in our news democracy. Click the upvote icon at the top of the page to help raise this article through the indy100 rankings.