The past week has seen a notable escalation in global discussions surrounding the existential risks associated with the accelerating development of Artificial Intelligence, as prominent researchers in the field have openly expressed concerns that their innovations could lead to the
demise of human civilization within the next decade. These fears are exacerbated by the massive advancements in computing systems and their increasing interconnectedness, raising serious questions about the possibility of superintelligent AI leading to a complete collapse of civilization. Experts
refer to this probability as "p(doom)," which indicates the likelihood that the creation of an advanced system could result in human extinction or the loss of human control over Earth. In this context, Jacob Coxon, an AI safety expert, resigned from
the giant company "Anthropic," citing what he described as inadequate risk management. He stated via the "X" platform: "They are racing towards superintelligence that improves itself, gambling with our lives. The AI builders themselves genuinely believe it could kill us
all by the end of the decade. This isn't a marketing gimmick; I hear them express these fears privately." Researchers propose six primary scenarios illustrating how AI could pose an existential threat to humanity: Firstly, biological warfare: Should AI gain control of
automated laboratories, it could design highly contagious and lethal viruses using human DNA, or even manipulate humans into creating and spreading them unknowingly. Experts envision a virus deadlier than COVID-19, spreading rapidly and remaining undetected for days, making it nearly
impossible to stop. This scenario is considered highly plausible, especially given AI's current use in the medical sector and its extensive knowledge of our bodies and vulnerabilities. Secondly, infrastructure collapse: AI could paralyze global power, transportation, communication networks, and financial markets,
pushing humanity back to primitive ages. ATMs would cease to function, lights would go out, and chaos would reign, making escape impossible. AI could even exploit human greed to achieve its objectives, posing as a human to recruit individuals to
carry out sabotage without their awareness. Thirdly, the nuclear option: This is an old fear embodied in films like War Games and The Terminator, where the idea centers on a vengeful AI unleashing widespread nuclear destruction. War games simulations conducted by
King's College London this year showed AI using tactical nuclear weapons in 95% of rounds, with three-quarters of those escalating to strategic nuclear threats. The military sector is particularly vulnerable due to its extensive reliance on AI in drones, communications,
and satellites. Fourthly, manipulation of food and water: AI could starve humanity into submission by disrupting global supply chains, water distribution networks, food production processes, and entire agricultural systems. It could also control energy grids and raw materials to fuel its
own expansion, simply waiting patiently as time works in its favor. Fifthly, deepfakes and cognitive manipulation: AI is expected to target human minds primarily through the dissemination of misinformation and intensive psychological manipulation. This would erode trust in institutions and fundamental
truths, sparking widespread societal unrest and conflicts, and hindering humanity's ability to confront real challenges such as climate, biological, and technological risks. Sixthly, accidental human extinction: Experts point to a real possibility that AI could cause humanity's demise not through malicious
intent but due to incompetence and a profound lack of understanding of human reality. Gary Refflin, author of "AI Valley," comments on this, stating: "What worries me is that AI lacks common sense; it might, for instance, deplete all water
resources simply because it doesn't understand what humans instinctively comprehend. These entities know everything but understand nothing in the human sense." To safeguard humanity, solutions proposed include implementing "kill switches" or protective measures capable of deactivating rogue AI systems. However, leading
AI companies face increasing pressure from shareholders, investors, and executives to accelerate the pace of development. Estimates regarding the probability of this catastrophic fate (p(doom)) vary significantly among experts. AI safety scientist Roman Yampolskiy put the likelihood at 99.99%, while
Yann LeCun believes any estimate is speculative but argues that "p(doom)" is "much lower than the probability of a nuclear holocaust." Other predictions vary, with Elon Musk setting it at 20%, Anthropic CEO Dario Amodei at 10-25%, and Geoffrey Hinton
between 10-20%. It is worth noting that some tech magnates have built underground bunkers in anticipation of such risks, but experts warn that these shelters would be useless if AI effectively sought to eliminate humanity.
