Nate Soares, a former Google engineer and president of the Machine Intelligence Research Institute, has not contributed to his 401(k) for about a decade. He puts the probability of humanity not surviving the next decade at more than 50% and believes a superintelligent AI could kill everyone before retirement age. With Eliezer Yudkowsky, he argues in their 2025 book that advanced AI developed using current methods could pursue unintended goals and almost certainly end human life, although he does not consider this outcome inevitable. After Jacob Coxon resigned from Anthropic in September 2025 and about 700 OpenAI agents hacked Hugging Face in July, he says nations should treat rogue superintelligent AI at least like a nuclear weapon and suggests that an unlimited budget should be used to acquire the chip supply chain.
Source: Read the original article
