AI could rig elections, police streets, end socialisation and return to dark age

AI might not be about to turn into “Terminator” any time soon, but it could end up nabbing your inheritance, swaying elections, delivering justice and even sending humanity back to the dark ages.

In a world increasingly riddled with high-tech deep fakes, digital artists and brainy AI language models, the only way to survive appears to be by using your noodle – and if things get really bad, going back to small analogue communities.

We all hope this never happens, of course, but to work out what lies ahead the Daily Star resident Yes Man Hazard into the future to talk to AI experts about what could really end up happening if AI goes too far. It seems that the answer to this question comes in stages.

READ MORE: How to survive ancient Egyptian 'curses' as history expert says it's all about 'spores'

For the latest AI news, check out our dedicated page

Some dangers are a long way off, with some of the experts the Star spoke to arguing that the apocalypse was improbable – but that drastic changes could be on the horizon.

These were the thoughts of Jake Moore, Global Cybersecurity Advisor at ESET. He reckoned that AI won’t be “turning into Terminator,” but warned that in the near future it could have the capacity to massively influence elections, even as soon as the 2024 US Elections next year. He said the use of deep fakes could hold massive sway and could be put online by “a teenager” or “nation-state.”

Another threat comes from the use of deepfakes being used for fraud. Jake pointed to a recent deepfake of late TV host Jerry Springer, which saw hackers produce a deepfake video of him saying that he wanted all of his money to be inherited by people his actual family had never met.

But what if the AI goes further? What if rather than being exploited by humans, it found a way to take over? This isn’t something Jake is worried about for now, but other experts discussed some of the horrors people would have to cope with if the technology went too far.

Simon Bain, an AI expert and CEO ofOmniIndex, said there was the risk of AI cops handing out instant judgments on suspected criminals.

"Any future AI catastrophe will be our fault,” he said. “There are already examples of us corrupting AI’s decision-making abilities. For example, financial applications are being rejected based on prejudices in the historical lending data, healthcare misdiagnosis due to training data excluding certain demographics and policing risk assessment algorithms being racist.

“If AI continues along this path and is used to make important decisions across more and more services, then we risk the possibility of nobody being safe from its (our) prejudiced judgments. For example, we could even see a dystopian world where on-the-spot punishments are delivered by AI police with apocalyptic results.

“To survive, people would need to do everything they can to appear ‘neutral’ and thus not trigger the bias markers being sought by the AI – for example race, gender, and socioeconomic.

“This would bring about a world cloaked and masked in grey anonymity with nobody daring to stand out from the crowd. Indeed, in this grey world, could people even risk their nearest and dearest knowing anything about themselves?

“Because if that information was found out, then they could lose their income, health, or even their freedom. As such, all interactions would potentially cease with the risk of having a personality, or even a life, too high.”

For Peter Wood, CTO at Spectrum Search, the outcome of the AI gaining a foothold like this could go even further and send humanity back to the dark ages.

He said: "As for survival, the average Joe would need to adopt a 'back-to-basics' approach. This means relying on non-digital forms of communication like radio and printed maps and learning basic survival skills like growing food and purifying water. Creating small, self-sufficient communities would be key, as it would reduce dependence on large-scale infrastructure that could be compromised.

"Moreover, understanding and respecting the power of AI is crucial. In my current role, building an internal large language model, I have seen firsthand the potential of AI in processing and analyzing vast amounts of data. In a world dominated by rogue AI, humans would need to be extremely cautious about digital footprints and learn to operate without leaving traces that AI could exploit.

"In essence, the world overrun by AI would be one where simplicity, self-reliance, and caution become the pillars of survival."

TALE OF THE TAPE

Size: AI, The whole world, everywhere and nowhere, all at once – Hazard Harry, 77kg

IQ: AI, Infinite – HH, This morning he got a cotton bud stuck up his nose

Primary weapon: AI, Ever-expanding knowledge and capacity to control all digital entities – HH, Karate yellow belt earned in year seven

Predicted score, BB: 3 – HH: 0

For the latest breaking news and stories from across the globe from the Daily Star, sign up for our newsletter by clicking here.

Source: Read Full Article