Six ways AI could make humans extinct in a horror digital doomsday

Six ways AI could make humans extinct in a horror digital doomsday Superintelligent AI poses a genuine existential risk to humanity within the decade, warn experts. Here's how it would do it... AI researchers sent shockwaves this week when they publicly admitted that their creations could wipe out humanity within the next decade. This digital apocalypse has become increasingly plausible as computers become more powerful and interconnected, with experts warning that a super-intelligent AI could turn rogue and destroy civilisation. They call it p(doom) - the probability that developing such an advanced system would result in human extinction or the complete loss of human control over the world. Jacob Coxon, an AI safety expert, resigned from his position at industry giant Anthropic in protest at what he views as a mishandling of the threat posed by the major players in the AI industry. He wrote on X: "They are racing straight to self-improving superintelligence and gambling with our lives." "The people building AI earnestly believe that it could kill us all by the end of the decade. This is not a marketing stunt. If anything, many executives and senior researchers will couch their phrasing in the press to sound sensible, but I hear the same people express fear privately. No other human activity poses this level of danger." However, if a supercomputer did decide that the planet would be better off without us, how would it go about it? Biological Warfare If 'Skynet' seizes control of automated 'cloud' laboratories, it could weaponise our own DNA against us by engineering highly contagious, deadly viruses capable of overwhelming healthcare systems across the globe and wiping out a significant portion of humanity in the process. Using the same virtual labs, AI could even manipulate us into creating and distributing the lethal pathogen ourselves. A supercharged COVID-19 that spreads rapidly yet remains undetected for days would be virtually impossible to contain. This scenario is far from far-fetched, as AI is already being deployed in the medical sector to develop treatments for illnesses. It already holds vast knowledge about our bodies and their weaknesses. Infrastructure AI could decide to strike at the very heart of where we live and work. We've been automating jobs for decades, almost all of which are controlled via the internet — precisely where an artificial superintelligence would operate most freely. Our digital adversary could effortlessly plunge us back into the Stone Age by disabling electrical grids, transport and communication networks, and financial markets. Put simply, the cash machines won't work, the lights will go out, and escaping the ensuing chaos will be near impossible. The AI could even exploit human greed to further its agenda, posing as a person to recruit unsuspecting individuals for acts of sabotage. Some of us could be unwittingly working for HAL without even realising it. The nuclear option A long-held fear and a sci-fi staple. From War Games to The Terminator, it's long been imagined that a vengeful AI could spell our doom by seizing control of our nuclear arsenal. Whether by launching the West or East against the other and waiting for the inevitable response, or by firing both sides' missiles simultaneously. In war games run by King's College London earlier this year, AI was quite gung-ho on using nuclear weapons. Some 95% of the games saw tactical nuclear use, and three-quarters reached strategic nuclear threats. The military as a whole is especially vulnerable because it already uses AI extensively - be that in drones, communications, or satellites. The Russian government has already boasted that it used drones guided entirely by a computer. Food and water A rogue artificial intelligence might simply starve us out. It could effortlessly disrupt global supply chains, water networks, food production, and agricultural systems. Such a system might commandeer the Earth's energy grid and raw materials for its own expansion. Essentially, the machine could simply bide its time - it has time on its side. Deep fakes It's conceivable that AI could target our minds before it targets our bodies. It could weaponise deepfakes, fake news and psychological manipulation to steadily erode public trust in institutions and established facts. This could lead to the collapse of the social foundations necessary for human society to function, resulting in widespread civil unrest, conflict, and undermining our ability to address climate, biological, or technological threats. By mistake It is conceivable that computers could wipe out humanity through incompetence rather than malice. Gary Rivlin, author of 'AI Valley' told The NY Post: "I worry about AI having no common sense. Then it somehow, stupidly, drains all the water because it doesn't understand what a human would. These things know everything, and they understand nothing." What are the chances? One way for humans to remain safe is to build kill switches or safeguards that could halt rogue AI agents in their tracks. However, leading AI companies are facing pressure from shareholders, investors, and their bosses to accelerate development, raising fears that safeguards are being overlooked. AI safety scientist Roman Yampolskiy has put the chances at 99.99%. AMI Labs founder Yann LeCun says any guess is a wild one, but p(doom) is "a lot less likely than a nuclear Holocaust". Other estimates are similarly varied: Elon Musk has put his p(doom) at around 20%, while Anthropic CEO Dario Amodei has estimated a 10%-25% chance of a catastrophic outcome. More recently, AI godfather Geoffrey Hinton put the figure at between 10% and 20%. Interestingly, some tech billionaires have constructed bunkers to protect themselves. However, experts warn that if AI decides to eliminate humanity, a bunker won't save anyone.