← the scenic garden rationality · Bayes · x-risk

Horizon & Prophecy · a network scenius · 2009–present

LessWrong & the rationalist diaspora

An online community devoted to human rationality and AI risk that spun off effective altruism, AI-safety research, and a Bay Area subculture.

WHERE Berkeley / online WHEN 2009 → present FORM the blog · the Sequences · the group house
Epistemic status: reported, medium confidence. Coded from public web sources (Wikipedia-led); not yet independently verified. Update your priors accordingly.
01

The idea-forge

LessWrong is the doctrinal home and idea-forge of the 21st-century rationalist subculture: a movement, seeded by Eliezer Yudkowsky's 2006–2009 Sequences on the group blog Overcoming Bias (run with Robin Hanson), that treats cognitive-bias correction and Bayesian probability1 as tools for both individual self-improvement and civilizational survival — given its core anxiety that badly-aligned artificial intelligence could end the species.

From a single blog it grew into forums, group houses, nonprofits, and a Bay Area-centered diaspora of successor projects.

02

The thesis they share

Human reasoning is systematically biased but improvable; by training in Bayesian rationality we can think more truthfully and act more effectively — above all to navigate the existential risk posed by advanced AI. — the unifying principle

03

Recurring motifs

  • The art of human rationality and overcoming bias
  • Bayesian reasoning as a way of life
  • AI alignment and existential risk
  • The Sequences and the group house
04

Why it caught

The promise of systematically winning by thinking better was flattering and habit-forming, creating devoted autodidacts. Yudkowsky's Sequences functioned as an addictive, identity-forming shared canon.2 And the framing of AI as an imminent existential stake turned an intellectual hobby into an urgent mission.

05

Internal dynamics

An online community anchored by a canonical text and forum, extending into a Bay Area subculture of meetups and group houses — intensely self-analytical and prone to schism.

06

The forces that gather & split

Convergent engines

  • Yudkowsky's Sequences & the LessWrong forum as shared canon and hub
  • Bay Area meetups and group houses
  • A common focus on rationality and AI risk

Divergent engines

  • Schisms with the "post-rationalist" diaspora
  • Tension between the rationality project and the AI-risk mission
  • Controversies and fraught overlaps (EA, tech elites)
07

What it did to the world

Spun off effective altruism, catalyzed the AI-safety field now central to frontier AI labs and policy, and shaped a distinctive Bay Area intellectual subculture — and its vocabulary.

A canonical text plus a forum and group houses that turned an online rationality project into a real-world subculture with outsized influence on AI. — the scenius signature

08

The community & its scribes

Two figures anchor the canon — Eliezer Yudkowsky, whose Sequences and Harry Potter and the Methods of Rationality became a recruitment funnel, and Scott Alexander (Slate Star Codex / Astral Codex Ten), whose blog carried the sensibility to a far wider readership while writing under a pen name — around a diffuse third author: the rationalist community itself.

Portrait of Eliezer Yudkowsky, photographed at Stanford in 2006.
Eliezer Yudkowsky, 2006 Researcher into artificial intelligence; author of the Sequences, photographed at Stanford. Photo © "null0" · Wikimedia · CC BY-SA 2.0
Eliezer Yudkowsky in 2008, wearing a cap reading "Bayesian Conspiracy."
Eliezer Yudkowsky, 2008 At the height of the Overcoming Bias / early LessWrong era — the cap reads Bayesian Conspiracy, the movement's in-joke for itself. Photo © Anders Sandberg · Wikimedia · CC BY 2.0
Scott Alexander The blogger behind Slate Star Codex and Astral Codex Ten — long pseudonymous, with no widely-verified free portrait. The absence is itself in-character. Pen name; no Creative-Commons likeness confirmed. Left uncredited by design rather than mislabel a namesake.
09

Mediums it worked in

medium / the blog · the forum medium / the Sequences medium / the meetup medium / the group house
10

Fate

Living and consequential: its once-fringe AI-risk concern moved to the center of the AI world, while the community keeps fissioning into successors.

Notes & provenance

  1. 1. Bayes' theorem — P(H|E) = P(E|H)·P(H) / P(E) — is the movement's mathematical touchstone for revising belief in light of evidence.
  2. 2. The Sequences (2006–2009) and Harry Potter and the Methods of Rationality served as the canon and the recruitment funnel that drew readers into the forum and, eventually, the Bay Area houses.
Coded from public web sources — provenance imported, confidence medium. Primary sources: "LessWrong," Wikipedia; "Rationalist community," Wikipedia; "History of Less Wrong," AlignmentForum; "Why Are There So Many Rationalist Cults?", Asterisk Magazine.
Images. All portraits are Creative Commons / public-domain from Wikimedia Commons, credited individually beside each plate: Yudkowsky (Stanford 2006) © "null0", CC BY-SA 2.0; Yudkowsky 2008 © Anders Sandberg, CC BY 2.0; Yudkowsky at Singularity Summit © David Orban, CC BY 2.0; Scott Alexander 2014 © Minda Haas Kuhlmann / UCinternational, CC BY 2.0; Scott Alexander & Larry Karaszewski © Poorlarry, CC BY-SA 4.0.