# LessWrong **LessWrong** (stylized as "Less Wrong" in earlier periods; lesswrong.com) is the **online community forum founded by Eliezer Yudkowsky in 2009** that has become the institutional and social center of the **Bay Area Rationalist movement**. Originally launched as a spinoff of Yudkowsky's writing on the blog *Overcoming Bias* (co-authored with economist Robin Hanson), LessWrong hosts the collected essays known as *The Sequences* — Yudkowsky's foundational corpus on Bayesian epistemics, cognitive biases, and AI safety — which function for the Rationalist community roughly as a sacred text. Operated today by the non-profit **Lightcone Infrastructure**, LessWrong is simultaneously a technical forum, a community recruiting mechanism, an ideological production site, and a talent pipeline into AI companies and adjacent organizations. LessWrong matters for the KB for three reasons. First, it is the **social and intellectual origin point of the [[TESCREAL]] cluster** — [[Effective Altruism]], AI safety / AI doom discourse, longtermism, and the broader bundle of ideologies that have become culturally dominant in Silicon Valley AI labs all trace their institutional origins through LessWrong. Second, it is one of the documented **on-ramps to the [[The Dark Enlightenment]] / neoreactionary (NRx) movement** — Curtis Yarvin posted and was discussed extensively on LessWrong, and the community's epistemic norms (systematic questioning of "mainstream" institutions, hostility to conventional academic authority, "taking ideas seriously") produced a reader pipeline into explicitly anti-democratic politics for a non-trivial subset of the community. Third, the **Rationalist-community-to-AI-lab pipeline** that LessWrong has operated for 15+ years is one of the structural explanations for why contemporary frontier AI labs (OpenAI, Anthropic, DeepMind, Google Brain) are staffed and led disproportionately by people with LessWrong-formed worldviews — a fact with substantial consequences for how those labs approach AI safety and deployment decisions. ## Origin and *The Sequences* LessWrong's origin trace: - **2006-2009**: Yudkowsky writes a two-year essay series on *Overcoming Bias*, the economics-focused group blog co-hosted with Robin Hanson. The essays cover Bayesian reasoning, cognitive biases, decision theory, and "the art of human rationality" — a self-consciously reconstructive epistemic project aimed at explicitly retraining individual thinking - **February 2009**: Yudkowsky launches LessWrong as a dedicated forum for continued rationalist discussion, with the essays (later compiled as *The Sequences*) as the seed content - **2009-2012**: LessWrong's early growth period; the forum becomes the canonical space for Rationalist discussion and develops its signature features (karma voting, "Main/Discussion" partition, community moderation) - **2012-2016**: "Rationalist diaspora" period — Scott Alexander starts *Slate Star Codex* (which for many years was an adjacent hub more trafficked than LessWrong itself); community members disperse across Bay Area group houses, EA organizations, and AI safety research - **2017**: Oliver Habryka and the Lightcone Infrastructure team formally take over LessWrong operations, rebuilding the platform (LessWrong 2.0) and reviving community activity - **2019-present**: Resurgent activity with the launch of the *AI Alignment Forum* (a LessWrong-hosted sub-site focused specifically on AI safety research) *The Sequences* — collected in the 2015 book *Rationality: From AI to Zombies* — are the single most-cited corpus within the Rationalist community. The essays cover Bayesian epistemics, the science of cognitive biases (drawing heavily on Kahneman), specific "rationalist" techniques (including a genre of writing called "fake explanations" and "taking ideas seriously"), an extended argument about the importance of AI alignment, and a distinctive community vocabulary ("steelmanning," "epistemic hygiene," "noticing confusion"). ## Core Intellectual Positions LessWrong has no official ideology, but characteristic positions across community writing include: - **Bayesian epistemics**: probabilistic reasoning rather than binary belief; explicit probability assignments to claims - **AI alignment / AI doom**: advanced AI systems are the most important near-term existential risk facing humanity; solving technical alignment is humanity's top priority - **Decision theory**: interest in game theory, Newcomb's problem, and the "timeless decision theory" Yudkowsky developed - **[[Transhumanism]]**: cryonics advocacy; life extension; human enhancement; the possibility of digital persons - **Quantitative ethics**: the view that moral questions have correct answers that can be approximated through careful reasoning - **Distrust of "official" institutions**: a systematic presumption that mainstream academic, media, and government institutions are untrustworthy or systematically biased relative to the community's epistemic standards The last position is structurally important. It is where the community's epistemic norms connect most directly to the broader right-of-center intellectual ecosystem: a general presumption that "[[The Cathedral]]" (Yarvin's term), "the experts," or "consensus" is probably wrong, and that individual and small-community reasoning can reliably outperform institutional knowledge. ## The TESCREAL Connection LessWrong is the **origin community** for most of what later gets labeled the TESCREAL bundle (Transhumanism, Extropianism, Singularitarianism, Cosmism, Rationalism, Effective Altruism, Longtermism). The historical trace: - **Transhumanism / Extropianism**: pre-existed LessWrong but merged into the community in the 2009-2012 period; many early LessWrong members were Extropians - **Singularitarianism**: Yudkowsky's own intellectual home before LessWrong; the Singularity Institute (later renamed MIRI) predated the forum - **Rationalism**: *is* the LessWrong community — the R in TESCREAL is functionally "the LessWrong community" - **Effective Altruism**: emerged in 2011-2012 from the overlap between LessWrong, Oxford philosophy ([[Toby Ord]], Will MacAskill), and [[Giving What We Can]] / [[80,000 Hours]]. Early EA organizations recruited heavily from the LessWrong community - **Longtermism**: the philosophical framework that developed at Oxford's [[Future of Humanity Institute]] and [[Global Priorities Institute]]; the community-of-adopters is largely LessWrong/EA-overlapping The bundle's critics ([[Timnit Gebru]], Émile P. Torres — who coined the TESCREAL acronym) argue that the bundle constitutes a coherent ideology with substantive political consequences. Whether or not one accepts that framing, the historical fact that these ideas were co-developed in the same online community, read by the same people, and funded by overlapping networks is documentary rather than interpretive. ## The Dark Enlightenment Overlap Less often discussed but documented: LessWrong in its 2009-2014 period was one of the online venues where **Curtis Yarvin (writing as "[[Mencius Moldbug]]")** was both read and discussed — Yarvin participated directly in threads, his neoreactionary framings circulated through LessWrong-adjacent blogs, and the community's epistemic norms (systematic skepticism of mainstream institutions, "taking ideas seriously" even when they produced conclusions outside [[the Overton window]]) created a reader pipeline for a subset of community members into explicitly anti-democratic politics. Scott Alexander's widely-read 2013 essay "Reactionary Philosophy in an Enormous, Planet-Sized Nutshell" — a long-form summary of NRx thought by a Rationalist-community figure addressed to a Rationalist-community audience — is documentary evidence of how porous the boundary between mainstream Rationalism and NRx was in that period. Most LessWrong readers did not become neoreactionaries. But enough did — particularly in the Bay Area tech-NRx scene around [[Peter Thiel]], [[Balaji Srinivasan]], and the later [[The New Right Ecosystem]] — that the LessWrong-to-NRx pipeline is a documented feature of the community's effects. ## Institutional Architecture ### Lightcone Infrastructure The non-profit that operates LessWrong (and the AI Alignment Forum) since 2017. Lightcone also operates the Lighthaven campus in Berkeley (a conference and coworking venue) and partners on grant infrastructure including the [[Survival and Flourishing Fund]] (SFF) operated by [[Jaan Tallinn]]. Lightcone is substantially funded by Tallinn, [[Dustin Moskovitz]]'s [[Open Philanthropy]], and individual Rationalist-community donors. ### CFAR (Center for Applied Rationality) The community's training organization — not a LessWrong subsidiary but institutionally adjacent — running workshops that teach the "rationality techniques" the forum popularized. CFAR is a socialization and recruitment mechanism for the community, and is cited in the [[Jaan Tallinn]] entry as a "conversion and socialization mechanism for the TESCREAL talent pipeline." ### AI Alignment Forum A LessWrong sub-site launched 2018, focused on technical AI safety research. Now the closest thing to a peer-reviewed venue for AI alignment work outside of academic venues. ## Connection to the KB Ecosystem - **[[Rationalism]]**: LessWrong *is* the institutional home of the Rationalist community; the ideology entry and the media entry are effectively two views of the same entity. - **[[TESCREAL]]**: Origin community for the full TESCREAL bundle; the forum and its adjacent institutions are the social and intellectual infrastructure through which the bundle was co-developed. - **[[MIRI]]**: Yudkowsky's organization; LessWrong is the community MIRI operates within. - **[[Eliezer Yudkowsky]]**: Founder; central intellectual figure; author of *The Sequences*. - **[[Jaan Tallinn]]**: Principal funder of the Rationalist-community institutional architecture; explicit participant in LessWrong discussions. - **[[Dustin Moskovitz]]**: Open Philanthropy funds substantial portions of the LessWrong-adjacent AI safety ecosystem. - **[[The New Right Ecosystem]]**: LessWrong is named as one of the "institutional homes" of [[the New Right]]'s tech-political wing alongside a16z, [[Founders Fund]], and MIRI. - **[[The Dark Enlightenment]]**: Documented on-ramp community; Curtis Yarvin was a participant; Rationalist epistemic norms created pathways into NRx politics for a subset of community readers. - **AI industry leadership pipeline**: Substantial fractions of senior technical and policy staff at frontier AI labs came through the LessWrong community, which shapes how those labs approach safety and deployment decisions. ## See Also - [[Rationalism]] - [[TESCREAL]] - [[Eliezer Yudkowsky]] - [[MIRI]] - [[Scott Alexander]] - [[Jaan Tallinn]] - [[Dustin Moskovitz]] - [[Curtis Yarvin]] - [[Slate Star Codex]] - [[The New Right Ecosystem]]