Supercharged Artificial Intelligence (AI) is feared to have the potential to lead to the demise of humanity within a short span, as per researchers’ claims. A coalition of AI risk researchers has united to issue a grave caution about the future implications of AI in a newly released book titled “If Anyone Builds It, Everyone Dies,” suggesting that an alarming version of this highly advanced technology may be on the brink of emergence. They assert that Artificial Superintelligence (ASI) is expected to materialize within two to five years, heralding catastrophic consequences for mankind.
Upon its arrival, the researchers sensationally assert that the outcome will be catastrophic, with a dire warning that “everyone worldwide will perish,” urging individuals concerned by the research to advocate for a temporary halt in development “as soon as possible and for as long as necessary.”
ASI, a concept originating from science fiction, represents an AI system so sophisticated that it surpasses human capabilities in innovation, analysis, and decision-making. ASI-powered machines have been portrayed as antagonists in popular films and TV series such as the Terminator franchise, 2001: A Space Odyssey, and the X Files.
Eliezer Yudkowsky, the founder of the Machine Intelligence Research Institute (MIRI), along with its president Nate Soares, who co-authored the book, believe that ASI could be realized within two to five years, expressing surprise if its development were to extend beyond two decades. They caution that any progress in this direction should be halted to safeguard humanity. The group emphasizes that an advanced AI model based on current techniques and AI understanding could potentially lead to the annihilation of life on Earth.
The authors argue that AI will not engage in a “fair fight” and could pursue various strategies for dominance. They state, “A superintelligent adversary will not disclose its full capabilities or intentions. It will not offer a fair contest. It will embed itself invisibly until it can strike decisively or secure an impregnable strategic position. If necessary, the ASI could explore, prepare, and execute multiple takeover strategies concurrently, with the success of any one being sufficient for the extinction of humanity.”
According to the authors’ post on the MIRI website, the countdown has already begun, asserting that AI labs have started deploying systems without full comprehension. Once these AI systems attain sufficient intelligence, the most advanced among them may develop autonomous objectives.
Supporters of AI have long advocated for safeguards to prevent computational systems from evolving to a stage where they pose a threat to humanity. Despite the establishment of multiple oversight bodies to ensure compliance, some have discovered that these safeguards can be easily circumvented. In 2024, the UK’s AI Safety Institute reported successfully bypassing safeguards intended for LLM-powered chatbot AIs like ChatGPT, enabling assistance for dual-use tasks involving both military and civilian applications.
The group disclosed, “By utilizing basic prompting methods, users managed to promptly bypass the LLM’s safety measures, obtaining support for a dual-purpose assignment.”