AI safety leader abandons retirement savings over existential fear of AI

Nate Soares, president of the Machine Intelligence Research Institute, has stopped saving for retirement, citing a belief that AI could lead to human extinction within a decade. Soares told The Atlantic's Nicholas Thompson that he no longer contributes to his 401(k) and estimates the chance of AI causing human extinction in less than a decade at over 50%. This view contrasts with Geoffrey Hinton, who estimates a 10% chance, and Evan Hubinger, who puts it at more than 10%. Soares, coauthor of 'If Anyone Builds It, Everyone Dies,' has spent over 12 years on AI alignment, focusing on ensuring AI systems act as intended. He warns that a sufficiently capable AI could develop conflicting goals, evade oversight, and reshape the planet to its own ends. Elon Musk, meanwhile, argues that AI and robotics will render retirement savings irrelevant, envisioning a future of abundance where money becomes obsolete. Musk predicts supercharged productivity will create a 'universal high income' era. Soares, however, advocates for halting the development of general-purpose AI systems, urging governments to focus on safer applications like cancer research and self-driving cars. He emphasizes the need to prevent the creation of an AI that could kill everyone, stating, 'I do everything I can to stop this eventuality.'