The Rationalist Movement: LessWrong and the AI Risk Debate

What Is the Rationalist Movement?

The rationalist movement is a community focused on improving human reasoning and reducing bias. Its members try to hold beliefs that match reality as closely as possible.

The movement is centered on the blog LessWrong. That site calls itself a community "devoted to refining the art of human rationality."

Rationalists study topics like logic, probability, and cognitive bias. They ask a simple question again and again: how do we think more clearly?

The word "rationalist" here has a special, modern meaning. It points to this specific community, not the old philosophy of pure reason.

More than a hobby

For many members, rationality is not just a hobby but a way of life. They apply it to careers, charity, and even how they run their homes.

The movement blends blogging, in-person meetups, and shared jargon. Insiders call themselves "the rationalists" or the "rationalist community."

Its ideas reach far beyond a niche blog today. They now shape how tech leaders and governments talk about artificial intelligence.

LessWrong and "the Sequences"

LessWrong is the main online home of the rationalist movement. AI researcher Eliezer Yudkowsky founded it in February 2009.

Yudkowsky seeded the new site with his own essays. Those essays are now known across the community as "the Sequences."

What the Sequences teach

The Sequences are a long series of blog posts about clear thinking. Yudkowsky wrote them between 2006 and 2009, first on the blog Overcoming Bias.

They describe how to avoid common failures of human reason. The goal is to think in ways that lead to true and accurate beliefs.

The essays cover probability, evidence, and self-deception. Many rationalists treat them as a shared starting text.

A hub for big ideas

LessWrong grew into the main online hub for the rationalist community. New members often read the Sequences before joining discussions.

LessWrong describes the Sequences as a foundation for several projects. These include the Machine Intelligence Research Institute and large parts of effective altruism.

The blog also became an early home for debate about AI danger. Yudkowsky used it to warn that smarter-than-human machines could be deadly.

The Bay Area Rationalist Community

The rationalist movement is strongest in the San Francisco Bay Area. Berkeley, in particular, became its real-world capital.

Many rationalists moved there to be near each other and near tech jobs. Some live in shared houses and attend regular meetups.

An online idea goes offline

What began as an online forum grew into a physical scene. Members host dinners, discussion groups, and multi-day retreats.

The community overlaps with startups, AI labs, and charities. Many members work in software or research by day.

This tight-knit world has its own culture and language. Outsiders sometimes find its inside jokes and jargon hard to follow.

Strengths and strains

Close bonds can spread ideas fast and fund new projects quickly. But they can also create pressure to agree and follow leaders.

Critics say the scene can feel insular and intense. Supporters say it is simply a group that takes ideas seriously.

Either way, the Bay Area scene gave the rationalist movement real weight. It turned scattered readers into a working network of friends and coworkers.

CFAR: The Center for Applied Rationality

The Center for Applied Rationality (CFAR) is a Berkeley nonprofit that teaches thinking skills. It grew directly out of the LessWrong community.

CFAR was founded in 2012 by Julia Galef, Anna Salamon, Michael Smith, and Andrew Critch. It runs workshops on rationality and cognitive bias.

Turning theory into practice

CFAR tries to turn the Sequences' ideas into real habits. Its workshops draw on psychology, math, and behavioral economics.

Participants learn tools for making better decisions under uncertainty. The aim is to form beliefs about the world as accurately as possible.

Anna Salamon has served as CFAR's president since 2021. The group's roots trace back to rationality work at the Machine Intelligence Research Institute.

A pipeline to AI safety

CFAR's workshops became a recruiting ground for AI-safety work. Many attendees went on to careers focused on AI risk.

This link made CFAR more than a self-help program. It helped funnel talented young people toward the movement's core mission.

Scott Alexander and Slate Star Codex

Scott Alexander is a psychiatrist and blogger who became a rationalist star. His blog Slate Star Codex drew a large, devoted audience.

He wrote long essays on science, medicine, politics, and reason. "Scott Alexander" was a pen name that hid his full legal identity.

The New York Times clash

In June 2020, Alexander deleted his blog in protest. He said a New York Times reporter planned to publish his real name.

Alexander feared the exposure would harm his psychiatric patients and his safety. He also cited past death threats against him.

The dispute became a major story about privacy and the press. Many fans rallied to his defense online.

The Astral Codex Ten era

Alexander later returned to writing on the platform Substack. His new blog is called Astral Codex Ten.

This time he wrote under his real name, Scott Siskind. His work remains a key voice in the rationalist world.

His readership includes engineers, founders, and researchers. That reach helped spread rationalist ideas into Silicon Valley.

How Rationalism Connects to Effective Altruism

Rationalism and effective altruism grew up side by side and overlap heavily. Many people belong to both communities at once.

Effective altruism asks how to do the most good with limited money and time. Rationalism asks how to think clearly in the first place.

Shared roots and members

LessWrong describes the Sequences as a base for much of effective altruism. The two movements share writers, funders, and events.

Both value evidence, careful reasoning, and long-term thinking. Both also worry a great deal about risks to humanity's future.

Money and talent flow easily between the two. A person may read LessWrong, attend a CFAR workshop, then take an effective altruism job.

One worry above all

For both groups, advanced AI became a top concern. They came to see it as one of the biggest risks the world faces.

This shared fear helped bind the communities together. It also pushed both toward the fast-growing field of AI safety.

The Seedbed of the AI-Safety Movement

The rationalist movement is widely seen as the seedbed of the AI-safety movement. Many of today's AI-risk leaders came out of LessWrong.

Yudkowsky spent years warning that superintelligent AI could end humanity. His writing shaped a generation of researchers and AI doomers.

From blog posts to global policy

Ideas first debated on LessWrong now appear in real policy fights. Concepts like the "alignment problem" started in these online forums.

Members founded labs, charities, and research groups around AI risk. Their work now touches governments and major AI companies.

To see the wider network today, read our guide to who is fighting AI. The rationalists sit near the center of that map.

A sober note on controversy

The movement has faced real controversy as it has grown. In 2025 a small violent offshoot called the "Zizians" drew national headlines.

Federal prosecutors tied the group to several killings, according to reporting by PBS and others. Most rationalists condemned the group and its actions.

Observers stress that the Zizians were a tiny fringe. Still, the case raised hard questions about intense, closed communities.

The wider movement kept its focus on research and debate. Most rationalists argue that the case reflects a splinter group, not the mainstream.

Conclusion: Why the Rationalist Movement Matters

The rationalist movement began as an online quest for clearer thinking. It grew into a force that helped launch the modern AI-safety debate.

From LessWrong and the Sequences to CFAR and Astral Codex Ten, its reach is wide. Its ideas now shape how the world argues about advanced AI.

Want to follow the people and ideas driving that debate? Read our daily AI briefing to stay current.