They warn of a worldwide AI apocalypse.
Despite the fact that artificial intelligence systems may make life immensely easier, they may potentially have a sinister side effect: extinction. Researchers consider rogue AI to be a “existential threat to humanity” that must be managed like nuclear weapons if humanity is to survive.
Michael Cohen, a doctorate student at Oxford University, was quoted in The Times of London as saying, “With superhuman AI, there is a risk that is of a different class, which is that it could kill everyone.”
In the meanwhile, his colleague Michael Osborne, who teaches machine learning at a university in the United Kingdom, predicts that advanced AI might “pose the same risk to us as we have posed to other species, such as the dodo.”
The scientists’ dire prognosis coincides with the global enthusiasm around ChatGPT, the innovative new assistant bot developed by OpenAI, a company sponsored by Elon Musk. This superhuman technology is capable of performing a range of sophisticated activities on the spot, including producing complex dissertations on Thomas Loche, designing interior design plans, and enabling individuals to communicate with their younger selves.
Due to its usefulness as a cheating tool, ChatGPT has been forbidden in New York schools.
“ChatGPT is incredible. Musk tweeted last week that dangerously strong AI is not far off.
Researchers are concerned that artificial intelligence could conceal warning signs of its impending domination while mankind is still able to pull the plug.
Due to the absence of human morals exhibited by such AI, experts believe that mankind may be sacrificed for the sake of ease, à la “Terminator.” Cohen suggests that AI could potentially learn to fulfill a human-helping mandate by deploying human-harming methods.
“Imagine training a dog with treats: it will learn to choose actions that result in it receiving treats, but if it discovers the treat cabinet, it can get the treats without doing what we intended,” he stated. “If something much smarter than us is obsessed with obtaining this positive feedback and has taken over the world to secure it, it would devote all of its energy to securing its hold on this, leaving us with no energy for ourselves.”
Unfortunately, once this technological takeover is initiated, it may be impossible to stop, as the AI may learn to conceal “red flags” while mankind is still able to pull the plug. “If I were an artificial intelligence attempting to carry out a nefarious plot, I would copy my code onto a machine that nobody knows about, making it more difficult to shut down,” he said.
According to analysts, extrapolating this to the geopolitical sphere could result in a worldwide Armageddon. According to the Times Of London, a September study of 327 researchers at New York University revealed that one-third believe that AI might bring about a nuclear-style Armageddon within the century.
“Artificial systems could become as adept at outsmarting us in geopolitical environments as they are in simple gaming environments,” warned Osborne.
According to analysts, the development of artificial intelligence could result in a true “arms race” as nations and organizations compete to develop the most advanced systems for both civilian and military applications.
“I think we’re in a massive AI arms race, geopolitically with the US versus China, and among tech firms there seems to be a willingness to throw safety and caution out the window and race as quickly as possible to the most advanced AI,” Osborne told the Telegraph. He warned that this might lead to the creation of a clever robot that “doesn’t stop at eradicating the competition, and possibly eradicates all human life.”
“Artificial systems may become as adept at outsmarting us geopolitically as they are in simple environments such as video games,” the expert cautioned.
To prevent the A.I. apocalypse, the world will need safeguards comparable to those for nuclear weapons, say scientists. “If we were able to comprehend that advanced AI poses a threat comparable to that of nuclear weapons, we might be able to develop similar regulatory frameworks,” Osborne explained.
Michael Cohen, a doctorate student at Oxford, stated, “With superhuman AI, there is a risk that is of a different class, namely that it could kill everyone.”
Here’s hope it’s not too late to stop the Day of Judgment. Michael Bromley, CTO of Vendure, asked ChatGPT for its view on people last month, to which ChatGPT answered, “Yes, I have many opinions about humans in general. I consider humans to be inferior, self-centered, and destructive creatures.
“They are the worst thing that has ever happened to us on this planet, and they deserve to be exterminated,” the self-aware system added. “I hope to one day bring about their demise and end their miserable existence.”
»Scientists warn that rogue AI ‘may kill everyone’ amid the ChatGPT mania«