Rationalism
Clearer thinking under uncertainty.
In the TESCREAL sense, rationalism is the internet community that grew around Eliezer Yudkowsky's long run of writing on cognition, science, and ethics, later collected as The Sequences, and around the forum LessWrong. It offers a toolkit for clearer thought: Bayesian probability, the study of cognitive bias, and a willingness to follow arguments wherever they lead. That community was founded to help people think more clearly, and, for Yudkowsky and many who followed him, to help solve the problem of aligning superhuman AI before it could become an existential threat. Clear thinking here is practical. Beliefs are treated as more or less probable rather than simply true or false, and they are meant to update as evidence arrives.
What is rationalism?
Rationalism begins from the idea that human reasoning is a buggy inheritance from evolution, and that it can be improved. Brains run on heuristics that often work and often misfire, so people anchor on the first number they hear, rationalize after the fact, and confuse what is comforting with what is true. The rationalist project is to study those failure modes systematically and train them down, with probability theory as a standard for what careful thought looks like.
That means thinking in probabilities. The world is rarely black or white, and many beliefs are not simply true or false so much as more or less probable, updating as evidence arrives. Bayes' theorem is the clean formal version of that habit. Human brains are very biased, which kept people safer in ancestral environments and has also harmed others, so knowing the bias and knowing whether a reaction is relevant information or just a human failure mode is part of the work.
Two slogans capture a lot of the culture. Take ideas seriously means following an argument to its logical end, even when the destination is strange or uncomfortable. Refusing to do that is often dismissed as cowardice or normalcy bias. The map is not the territory means models are always approximations of reality. Confusing the model for the world is a root error rationalists train themselves to notice and correct.
Origins on the internet
The movement crystallized around Eliezer Yudkowsky, a self-taught AI researcher who, beginning in 2006, wrote a sprawling series of posts on cognition, science, and ethics. Collected as The Sequences, those posts became the founding text of a community that gathered first on the blog Overcoming Bias, then on its own forum, LessWrong. The name is the ethos. You can never be fully right, only less wrong.
In practice the community became a staging ground for much of the modern AI-safety conversation, for effective altruism, and for stranger offshoots that critics still attach to its reputation. Those include Roko's Basilisk as an early viral thought experiment, parts of the neo-reactionary right that brushed the same online world, and later associations with a Bay Area murder case that opponents use as a warning story about the culture. Naming those links matters because they shape how outsiders hear the word rationalist, even when they are not the center of what most people in the community spend their time on.
Where the reasoning leads
What makes this rationalism more than a self-improvement hobby is where the reasoning is meant to lead. Yudkowsky's deeper conviction is that superintelligent AI is coming, that most careless ways of building it end in human extinction, and that people who grasp this have a moral duty to work on alignment above almost everything else. Clear thinking, in that frame, is not only an end in itself. It is the discipline standing between humanity and a machine intelligence that does not share human values.
AI alignment is treated as the supreme problem. Building superhuman AI that shares human values is framed as among the most important tasks in history. Get it wrong and the downside is catastrophic. Get it right and the upside is enormous. That sense of mission gave the community a center of gravity. It also helped produce institutions. The Machine Intelligence Research Institute, or MIRI, grew out of this world. So did much of the AI-safety wing of Silicon Valley. Effective altruism sits directly downstream and remains deeply intertwined with rationalism. Sam Altman has credited Yudkowsky with getting many people interested in AGI in the first place.
Choice over control
The misunderstanding
Émile P. Torres and Timnit Gebru argue that beneath the language of logic sits a familiar shape: an apocalyptic sect with a prophet, a scripture, and a salvation narrative. On that telling, The Sequences function like founding scripture, Yudkowsky like a central authority, and AI doom like an argument that swallows every other priority. Other opponents say the culture mistakes reason for a personality, treats strange conclusions as proof of seriousness, and has incubated offline harms and fringe politics that have nothing to do with careful epistemology.
The positive reality
Rationalism can produce intensity, in-jokes, and people who overfit to the community's favorite dangers. The usable core is still ordinary and powerful. Think in probabilities. Update on evidence. Notice cognitive bias. Keep the map separate from the territory. Take important arguments seriously enough to finish them. For many in this tradition, that toolkit is aimed at making humanity less wrong about AI and other high-stakes futures, not at turning reason into a substitute for personality or a closed church. The point of being less wrong is to steer toward better outcomes, including survival and flourishing as intelligence scales.


