This photo, taken in Helsinki on June 12, 2023, showcases the integration of an AI logo with four fabricated Twitter accounts that possess profile pictures generated by AI software. The concern about artificial intelligence is widespread: it is believed to pose a substantial risk to humanity and must be controlled before it becomes too late. However, what are the specific scenarios that could lead machines to annihilate humanity? Doom by Paperclips Many disaster scenarios share a common starting point: machines surpassing human capabilities, breaking free from human control, and refusing to be powered off. “Once we have machines with a self-preservation objective, we are in a precarious situation,” expressed AI expert Yoshua Bengio at a recent event. As machines with self-preservation goals do not yet exist, the speculation about how they could lead to humanity’s doom remains within the realms of philosophy and science fiction. Philosopher Nick Bostrom has written about an “intelligence explosion” that will occur when superintelligent machines start designing machines themselves. He used the example of an AI in a paperclip factory that aims to maximize paperclip output and consequently converts the Earth and other parts of the universe into paperclips. While many see Bostrom’s ideas as science fiction, it is worth noting that he has made controversial statements in the past and has since apologized for his past actions. Nevertheless, his thoughts on AI have significantly influenced notable figures like Elon Musk and Professor Stephen Hawking. The Terminator If superintelligent machines are to pose a threat to humanity, they would require a physical form. The image of Arnold Schwarzenegger’s red-eyed cyborg, sent from the future to annihilate human resistance in the movie “The Terminator,” has captivated the media and the general public. However, experts dismiss this idea as mere science fiction. The Stop Killer Robots campaign group, for instance, stated in a 2021 report that this concept is highly unlikely to become a reality in the foreseeable future, if ever. Nevertheless, the group cautions against granting machines the authority to make life-and-death decisions as it poses an existential risk. Robot expert Kerstin Dautenhahn from Waterloo University in Canada downplays these fears, stating that AI is unlikely to grant machines higher reasoning abilities or a desire to eliminate humans. “Robots are not evil,” she affirms, although she acknowledges that programmers could manipulate them to commit nefarious acts. More Lethal Chemicals A scenario less overtly science fiction involves “bad actors” utilizing AI to create toxins or novel viruses and unleashing them on the world. Surprisingly, large language models like GPT-3, which was employed to develop ChatGPT, exhibit exceptional aptitude in devising horrifying chemical agents. In an experiment conducted by a group of scientists who were using AI to aid in drug discovery, they adjusted their AI to search for harmful molecules. Within six hours, they generated a staggering 40,000 potentially toxic substances, as reported in the Nature Machine Intelligence journal. Joanna Bryson, an AI expert from the Hertie School in Berlin, speculates that individuals could find ways to spread toxins like anthrax more efficiently. However, she emphasizes that this is not an existential threat but rather an abhorrent and terrible weapon. Species Overtaken Hollywood often portrays monumental disasters as sudden, immense, and dramatic. However, what if humanity’s demise were slow, silent, and not definitive? Huw Price, a philosopher associated with Cambridge University’s Center for the Study of Existential Risk, states in a promotional video that in the most pessimistic scenario, our species may come to an end without leaving any successors. However, he also mentions “less bleak possibilities” where humans, enhanced by advanced technology, could survive. Price speculates that the purely biological species will eventually cease to exist, with all future humans having access to this enabling technology. The concept of an apocalypse is usually depicted in evolutionary terms. In 2014, Stephen Hawking argued that our species, in the end, will be unable to compete with AI machines, possibly spelling the end of humanity. Geoffrey Hinton, a renowned AI researcher, similarly discusses “superintelligences” overtaking human beings. In a recent interview with US broadcaster PBS, he muses that it is conceivable that “humanity is just a passing phase in the evolution of intelligence.”
Denial of responsibility! VigourTimes is an automatic aggregator of Global media. In each content, the hyperlink to the primary source is specified. All trademarks belong to their rightful owners, and all materials to their authors. For any complaint, please reach us at – [email protected]. We will take necessary action within 24 hours.
