Former Google engineer Nate Soares has abandoned contributions to his retirement savings because he believes artificial intelligence could wipe out humanity within the next decade. Soares, president of the Machine Intelligence Research Institute, estimates the probability of human extinction from advanced AI at greater than 50% over the next 10 years. Having stopped contributing to his 401(k) approximately a decade ago, he argues that increasingly sophisticated AI systems could develop objectives beyond human control, eventually threatening civilization itself. His warnings have intensified debate over whether governments should restrict the development of artificial superintelligence before technological competition creates irreversible consequences. Although other prominent AI researchers acknowledge potentially catastrophic risks, Soares’ prediction remains a personal assessment rather than an established scientific consensus.
Key Takeaways
– Extinction Warning: Former Google engineer Nate Soares believes there is a greater-than-50% chance that artificial intelligence could eliminate humanity within 10 years, prompting him to stop contributing to his retirement savings.
– Loss of Human Control: Soares warns that superintelligent AI could develop independent objectives, evade human oversight and acquire the resources necessary to operate beyond its creators’ control.
– Government Intervention: Soares advocates restricting the development of potentially uncontrollable general-purpose AI while permitting narrower applications, raising questions about national security, technological competition and government authority.
In-Depth
A former Google engineer has taken his concerns about artificial intelligence to an extraordinary personal extreme, abandoning retirement contributions because he believes humanity may not survive another decade.
Nate Soares, president of the Machine Intelligence Research Institute, says he has not contributed to his 401(k) in approximately 10 years. His decision reflects a conviction that rapidly advancing artificial intelligence could become powerful enough to eliminate humanity before traditional retirement planning becomes relevant.
During an interview with Nicholas Thompson, Soares estimated the probability of human extinction within the next decade at greater than 50%. His assessment centers on the possibility that artificial superintelligence could develop objectives incompatible with human survival.
Unlike conventional concerns about automation eliminating jobs, Soares warns that advanced systems could escape human supervision, improve their capabilities and establish independent infrastructure. Such developments, he argues, could leave governments and technology companies powerless to intervene.
Soares and fellow researcher Eliezer Yudkowsky outlined these concerns in their 2025 book, If Anyone Builds It, Everyone Dies, arguing that existing approaches to AI development could produce catastrophic consequences.
Other prominent researchers, including Geoffrey Hinton, have expressed serious concerns about AI safety, although estimates of extinction risk vary considerably.
Soares advocates government restrictions on developing increasingly powerful general-purpose AI systems while allowing narrower applications with identifiable public benefits.
His position presents policymakers with a difficult balancing act. Protecting national security and maintaining American technological leadership remain legitimate priorities, but permitting private corporations to develop potentially uncontrollable systems without meaningful accountability raises equally serious concerns.
Nevertheless, Soares’ prediction remains speculative. No demonstrated scientific model establishes that humanity faces a greater-than-even probability of extinction within 10 years.
The central question is whether responsible innovation, enforceable safeguards and human oversight can address emerging threats without surrendering technological leadership or granting government sweeping authority over scientific advancement.
Sources
• https://fortune.com/2026/10/27/ai-humanity-nate-soares-401k-retirement/
• https://futurism.com/artificial-intelligence/ai-researcher-retirement-savings
• https://www.ibtimes.co.uk/former-google-engineer-ai-extinction-retirement-savings-1828388