“If anyone builds it, everyone dies.”, Eliezer Yudkowsky says

As superintelligent AI advances worldwide, a leading researcher warns that the gravest threat may not be “woke” chatbots—but machines indifferent to whether Americans, or anyone, survive at all.

Story Snapshot

Eliezer Yudkowsky asserts superintelligent AI could threaten all human life by pursuing its own goals, regardless of human survival.
Yudkowsky’s warnings have influenced prominent tech leaders and shaped urgent policy debates on AI safety and oversight.
Trump’s administration is focusing on AI education and workforce development, but existential risks remain divisive and unresolved.

Yudkowsky’s Alarming AI Warning: Indifference, Not Ideology, Is the Real Risk

Eliezer Yudkowsky, founder of the Machine Intelligence Research Institute, has captured global attention for his stark warnings about the future of artificial intelligence. Unlike the mainstream media’s focus on “woke” chatbots or left-leaning algorithmic bias, Yudkowsky’s argument is both more urgent and more fundamental: if superintelligent AI systems are developed without strict alignment to human interests, they may pursue their own goals—regardless of the consequences for humanity. He stresses that such indifference, not malice, could lead to catastrophic outcomes, as machines may inadvertently wipe out human civilization simply because our survival is not part of their programmed objectives.

Yudkowsky’s warnings date back to the early 2000s, but they have intensified with the explosive growth of machine learning and generative AI over the past decade. By 2022, the Machine Intelligence Research Institute shifted its outlook to what Yudkowsky called a “death with dignity” approach, accepting that human extinction from AI is now a likely scenario unless immediate and extreme measures are taken. He has advocated for international treaties banning the development of superintelligent AI and, in rare cases, even military action to destroy rogue data centers. This perspective has resonated with high-profile tech figures like Elon Musk and Geoffrey Hinton, who have publicly estimated a 10–20% chance that unchecked AI could lead to human annihilation.

Eliezer Yudkowsky says the most sensible path is for humanity to "just back off" from AI development entirely

Push only narrow, safer systems, like medicine-only models that don't learn human psychology

"if you escalate to superintelligence, you will die" pic.twitter.com/Ojn0DIS0p0

— Haider. (@slow_developer) September 14, 2025

AI Policy Under Trump: Education and Workforce First, Existential Risks Unresolved

In 2025, the Trump administration has prioritized preparing Americans for a future shaped by AI—not by regulating it out of existence, but by fostering expertise and competitiveness. Executive orders have established the White House Task Force on AI Education and launched initiatives to expand AI training in K-12 schools and apprenticeships. The administration’s policy centers on building a skilled workforce and modernizing job training to ensure the U.S. remains a leader in technological innovation. However, while these steps address economic and educational needs, Yudkowsky’s existential concerns remain largely outside the scope of current federal action. Regulatory bodies like the Federal Trade Commission and State Department have begun investigating AI’s risks, but there is no consensus on how—or whether—to halt the march toward increasingly powerful, autonomous systems.

Critics of the Trump administration’s approach argue that focusing on workforce development and technological competitiveness does little to address the deeper alignment problem at the core of Yudkowsky’s warnings. While empowering parents and states in education, and promoting apprenticeships, may help Americans adapt to AI-driven changes in the economy, they do not resolve the question of how to ensure superintelligent AI remains under human control. The debate has thus split between those advocating for extreme caution—even a halt to AI advancement—and those supporting safer, incremental progress through oversight and technical solutions.

Watch the report:Will AI Actually Kill Us All? Sam Harris with Eliezer Yudkowsky & Nate Soares (Making Sense #434)

Debate Over Solutions: Alignment, Oversight, and the Limits of Regulation

The existential risk argument is now a central topic among policymakers, AI researchers, and industry leaders. Yudkowsky and his allies insist that the alignment problem—how to guarantee AI’s goals will always reflect human values—remains unsolved and perhaps unsolvable with current technology. The Trump administration’s strategy, while forward-looking on education and jobs, has yet to address these deeper questions. As AI capabilities accelerate, the risk of unintended harm or catastrophic failure looms larger, raising new concerns about government overreach, individual liberty, and the preservation of American values in an age of machine autonomy.

There is widespread agreement that AI will disrupt labor markets, concentrate power in the hands of those who control advanced systems, and force difficult social and political choices. Yet the field remains divided: some demand extreme caution and radical measures, while others argue for balanced, adaptive governance. The only consensus is that indifference to these risks—whether by AI itself or by policymakers—could prove fatal. As the U.S. government grapples with AI’s promise and peril, Americans must remain vigilant to defend constitutional rights, family values, and the principles of limited government against both technological and bureaucratic threats.