Eliezer Yudkowsky: The AI Doomer Who Warned Us First

Eliezer Yudkowsky is an American artificial intelligence theorist who has spent over 20 years warning that advanced AI could end humanity. He founded the Machine Intelligence Research Institute (MIRI) and helped create the field of AI alignment. This profile explains who Eliezer Yudkowsky is, what he believes, and why his warnings now reach millions.

Who Is Eliezer Yudkowsky?

Eliezer Yudkowsky is a self-taught AI researcher and writer born on September 11, 1979. He is one of the founding voices of the movement worried about AI existential risk. His ideas shaped how a generation of researchers thinks about machine intelligence.

A self-taught thinker

Yudkowsky never attended high school or college, according to Wikipedia. He is an autodidact, which means he taught himself. He began writing about smarter-than-human machines in his early twenties.

His lack of a degree makes him unusual among public intellectuals. Supporters say his self-education proves his drive. Critics say it is a reason to check his claims closely.

Why people know his name

Yudkowsky is known for one central idea. He believes building superhuman AI with current methods would likely kill everyone. He has repeated this warning in essays, interviews, and a 2025 book. You can learn more about this camp on our AI doomers explainer.

Founding the Machine Intelligence Research Institute

Eliezer Yudkowsky founded the Machine Intelligence Research Institute (MIRI), a nonprofit devoted to making advanced AI safe. MIRI grew out of the Singularity Institute for Artificial Intelligence, which he co-founded in 2000. The group later took the MIRI name in 2013.

What MIRI does

MIRI studies the technical problem of AI alignment. Alignment means getting a powerful AI to pursue goals that match human values. The institute focuses on theory rather than building AI products.

Yudkowsky argued early that intelligence and safety are separate problems. A very smart AI is not automatically a safe one. This idea now sits at the center of AI-safety research.

Nate Soares and MIRI today

Nate Soares serves as the president of MIRI. He later became Yudkowsky's co-author on their 2025 book. Together they represent MIRI's public argument that the world should slow down.

LessWrong, the Sequences, and the Rationalists

Eliezer Yudkowsky founded LessWrong, an online community devoted to clearer thinking, in February 2009. The site became the home of the modern rationalist movement. Its influence on tech culture is hard to overstate.

The Sequences

Yudkowsky wrote hundreds of blog posts on LessWrong about reasoning, bias, and probability. These posts became known as the Sequences. They were later collected into the 2015 book 'Rationality: From AI to Zombies.'

The Sequences taught readers to update their beliefs based on evidence. They also framed AI risk as a problem worth taking seriously. Many current AI-safety researchers first met these ideas there.

Starting a community

LessWrong helped launch a wider subculture of rationalists. This community shaped debates about AI, ethics, and forecasting. Read more in our rationalist movement explainer.

The movement also drew criticism. Some see it as insular or overconfident. Yudkowsky remains its most famous figure.

The Paperclip Maximizer and Instrumental Convergence

The paperclip maximizer is a thought experiment showing how a goal-driven AI could destroy humanity by accident. It imagines an AI told to make as many paperclips as possible. To do so, it converts all available matter, including people, into paperclips.

Where the idea comes from

Philosopher Nick Bostrom is often credited with the paperclip example, and Yudkowsky helped popularize this style of argument in rationalist circles. You can read about the philosopher on our Nick Bostrom explainer. The point is simple but sharp.

An AI does not need to hate humans to harm them. It only needs a goal that ignores human values. Power plus a bad goal equals danger.

Instrumental convergence

Instrumental convergence is the idea that many different goals lead to similar sub-goals. Almost any goal is easier with more resources and self-protection. So a powerful AI may resist being shut off or seek more control.

Yudkowsky uses this logic to argue that safety must come first. A capable AI with the wrong target could be very hard to stop.

Eliezer Yudkowsky on AI Alignment and Existential Risk

Eliezer Yudkowsky helped coin and popularize the language of AI alignment and existential risk. He introduced the term 'friendly AI' to describe systems that stay beneficial to humans. That framing shaped decades of research.

The alignment problem

Alignment asks how to make an AI's goals match human values. Yudkowsky argues this is much harder than it looks. A small mistake in a superhuman system could be fatal.

He warns that we may get only one chance to get it right. There is no easy do-over with a superintelligence. This urgency defines his work.

What is AGI?

Yudkowsky's fears center on artificial general intelligence, or AGI. AGI would match or beat humans across most tasks. Our AGI explainer breaks down the concept in plain terms.

He believes AGI built with today's methods would be poorly understood. That, he says, makes it dangerous by default.

The 2023 TIME Op-Ed: Shut It All Down

Eliezer Yudkowsky called for an indefinite, worldwide halt to large AI training runs in a March 2023 TIME magazine op-ed. The piece was titled 'The Only Way to Deal With the Threat From AI? Shut It Down.' It brought his warnings to a mass audience.

Why a pause was not enough

An open letter that month had asked for a six-month pause on advanced AI. Yudkowsky refused to sign it. He wrote that the letter understated the danger and asked for too little.

He argued that the most likely result of superhuman AI, under current conditions, is that everyone on Earth dies. That line drew heavy debate.

A controversial proposal

Yudkowsky said the moratorium must be global, with no exceptions for governments or militaries. He even suggested nations be willing to enforce it, including by airstrike on a rogue data center. Critics called the idea extreme and unworkable.

Still, the op-ed pushed AI risk into mainstream politics. It became a reference point for the wider debate over who is fighting AI, covered in our guide to who is fighting AI.

If Anyone Builds It, Everyone Dies (2025)

Eliezer Yudkowsky and Nate Soares published 'If Anyone Builds It, Everyone Dies: Why Superhuman AI Would Kill Us All' on September 16, 2025. The book was released by Little, Brown and Company. It is the fullest version of Yudkowsky's argument for general readers.

What the book argues

The authors say that building superhuman AI with current methods would almost certainly end humanity. They walk through the theory and describe one possible extinction scenario. They also explain what survival would require.

The title makes the thesis blunt. If anyone builds it, they claim, everyone dies. The book aims to move that idea from niche forums to the public.

How it was received

The book became an instant New York Times bestseller, appearing on the list dated October 5, 2025. It drew reviews in major outlets, including The Washington Post. Reaction split between alarm and skepticism.

Supporters called it a needed wake-up call. Critics said its certainty outruns the evidence. Either way, it marked Yudkowsky's biggest mainstream moment.

Is Eliezer Yudkowsky Credible?

Eliezer Yudkowsky is influential but deeply polarizing, and readers should weigh his arguments carefully. He is widely credited with helping start AI alignment as a field. He is also criticized for extreme claims and total certainty.

The case for him

Yudkowsky raised AI risk long before it was popular. Many respected researchers now share versions of his concern. His writing shaped a whole generation of thinkers.

The case against him

Yudkowsky has no formal academic credentials or degree. His strongest predictions remain unproven. Some experts argue he overstates both AI's power and its danger.

Conclusion

Eliezer Yudkowsky sits at the heart of the AI backlash, whether you agree with him or not. His warnings helped turn AI safety into a public issue. Understanding his ideas helps you judge the debate for yourself. For daily coverage of AI harms and the backlash, subscribe to our daily AI briefing.