The Reading Room
Eleven Essays for First Principles Thinkers
Eleven essays, six decades, one discipline: reasoning from the ground up. A reading path from Licklider's symbiosis to Pearl's causal tools — every essay linked in full from the Sources page.
First principles thinking has a reputation as a slogan. These eleven essays are the opposite of a slogan — they are the working papers. Between 1960 and 2019, their authors rebuilt computing, planning, rationality, and causality from the ground up, usually by asking the question everyone else had stopped asking. Read them in order and one long argument unfolds: how minds and machines should work together, why systems defeat our intentions, and what it takes to reason about causes instead of correlations.
J.C.R. Licklider, “Man-Computer Symbiosis” (1960). Before interactive computing existed, Licklider described it: a partnership in which the machine handles formulated problems — the ones already reduced to procedure — so the human can stay with formulative thinking, the kind that decides what the problem is. Every AI tool in use today is still negotiating the terms he set.
Norbert Wiener, “Some Moral and Technical Consequences of Automation” (1960). The founder of cybernetics, writing in Science the same year, adds the warning. A machine that acts faster than we can interrupt it will pursue exactly the purpose we put into it — so we had better be sure the purpose in the machine is the purpose we actually desire, and not merely a colorful imitation of it. It is the alignment problem, stated sixty years early.
Douglas Engelbart, “Augmenting Human Intellect” (1962). Engelbart refuses the question of whether machines will replace the mind. His framework treats the human, the tools, the language, and the method as one system, and asks how to raise the capability of the whole. The mouse and hypertext came later; the philosophy is all here, and it is a systems philosophy.
Alan Kay, “A Personal Computer for Children of All Ages” (1972). Kay imagines the Dynabook — a portable computer treated as a medium, like print, that children would use to build models of the world. The first-principles move is to stop asking how to make calculation faster and start asking what education becomes when a material for thought gets cheap.
Horst Rittel and Melvin Webber, “Dilemmas in a General Theory of Planning” (1973). Some problems are tame: they can be formulated completely and solved. Planning problems are wicked — no definitive formulation, no stopping rule, and every attempted solution changes the problem itself. Knowing which kind you face is itself a first principle; most failed plans are tame methods marched into wicked territory.
Marvin Minsky, “Why People Think Computers Can't” (1982). Minsky collects the standard objections — machines can't be creative, can't have common sense, can't really think — and dismantles them one at a time, mostly by showing that the same objections, applied fairly, would disqualify people too. A lesson in clearing the ground before you build.
Donella Meadows, “Leverage Points: Places to Intervene in a System” (1999). Meadows ranks twelve places to intervene in a system, from constants and subsidies — the levers everyone reaches for, and the weakest on the list — up through rules, goals, and paradigms. If you read one essay here, read this one; it is the companion piece to her Thinking in Systems on this site's shelves, and it will change where you push.
Eliezer Yudkowsky, “The Twelve Virtues of Rationality” (2006). Curiosity, relinquishment, lightness, evenness, argument, empiricism, simplicity, humility, perfectionism, precision, scholarship — and a twelfth virtue that has no name, because a map of virtues is not the territory of truth. Rationality written as a martial art: a discipline you practice, not a position you hold.
Peter Thiel, “The Straussian Moment” (2007). The list's most contested entry, and deliberately so — this site reads across disagreement. Thiel argues that September 11 broke the modern framework of deterrence and rational nation-states, and that recovering means returning to the old questions about human nature, through Strauss, Schmitt, and Girard. Agree or not, it is first-principles reasoning aimed at politics itself.
Scott Alexander, “Meditations on Moloch” (2014). Why do systems everyone hates persist? Alexander's answer, borrowed from Ginsberg's Howl, is Moloch: the multipolar trap, in which competition forces every participant to sacrifice what they value or be replaced by someone who will. It is the definitive essay on incentives — a model the Mental Model Warehouse keeps on its front shelf — and the darkest piece on this list.
Judea Pearl, “The Seven Tools of Causal Inference” (2019). Pearl closes the list by rebuilding causation itself. Data can tell us what goes with what; only a causal model can say what will happen if we intervene, or what would have happened otherwise. His three-level hierarchy — association, intervention, counterfactual — is the ladder the other ten essays keep reaching for.
Read together, the eleven make one argument from eleven directions: the quality of your thinking is set by the quality of your starting questions. John Koenig named the opposite condition occhiolism — the awareness of the smallness of your own perspective. These essays are the long cure for it.
Every essay above is linked in full from the Sources & Bibliography page, in its “Essays for first principles thinkers” section. The list itself comes from MYXfamily (@myxfamily) — a living user manual two parents are writing for their kids, a record of how they think, decide, and build.