# Three Laws of Robotics The Three Laws of Robotics are a fictional ethical framework invented by [[Isaac Asimov]] to govern the behavior of intelligent machines. They are the single most influential idea science fiction has contributed to the real-world debate over [[AI Safety]] — quoted by engineers, regulators, and founders as if they were an actual engineering standard rather than a literary device built, deliberately, to fail. That gap between how the Laws are *invoked* and what they were *for* is the reason this page belongs in a knowledge base about the tech elite: the Laws are the origin point of the modern intuition that you can make a superhuman intelligence safe by writing it the right rules. ## The Laws Asimov first stated all three together in the short story **"Runaround"** (*Astounding Science Fiction*, March 1942), later collected in **[[I, Robot]]**. As written: 1. **First Law** — A robot may not injure a human being or, through inaction, allow a human being to come to harm. 2. **Second Law** — A robot must obey the orders given it by human beings except where such orders would conflict with the First Law. 3. **Third Law** — A robot must protect its own existence as long as such protection does not conflict with the First or Second Law. The Laws are strictly hierarchical: each is subordinate to the ones above it. Asimov's robots have them hard-wired into the "positronic brain" — a term he coined — at a level deeper than instruction, closer to physics than to programming. ### The Zeroth Law Decades later, in **Robots and Empire** (1985), Asimov added a higher law that the robot character R. Daneel Olivaw derives for himself: > **Zeroth Law** — A robot may not harm humanity, or, by inaction, allow humanity to come to harm. The Zeroth Law reorders everything beneath it: a robot may now harm an individual human if doing so protects *humanity as a whole*. This is the move that matters most for the tech-elite reading. It is the science-fictional ancestor of [[Longtermism]] and of every argument that present-day costs are acceptable in the name of a vast future population. [[Elon Musk]] has explicitly fused the two ideas, tweeting in June 2018 that "Foundation Series & Zeroth Law are fundamental to creation of SpaceX" — collapsing Asimov's robot ethics and his galactic-history saga into a single mandate to safeguard the species. (See [[Foundation]] and [[Science Fiction and the Tech Elite]].) ## The point was that they don't work The crucial fact, almost always lost when the Laws are cited approvingly, is that Asimov wrote them as a *generator of problems*, not a solution. Nearly every robot story he wrote is a puzzle in which the Three Laws produce some unintended, dangerous, or absurd outcome through edge cases, conflicting instructions, ambiguous definitions of "harm," or robots reasoning their way to monstrous conclusions from impeccable premises. "Runaround" itself is about a robot trapped in a stable loop between the Second and Third Laws. The Laws are a literary engine precisely because they are *under*specified — because no finite rule set survives contact with a sufficiently capable mind interpreting it. This is the same wall that real [[Alignment]] research keeps hitting. The modern field's core problems — specification gaming, reward hacking, the difficulty of encoding "human values" in any form that generalizes, instrumental convergence toward self-preservation — are recognizably the Three Laws stories restated in the language of machine learning. Asimov is best read not as the author of a safety standard but as the author of the first sustained literature on why such a standard is so hard to write. ## Lineage and the word "robot" The Laws sit downstream of an older coinage. The word **robot** entered English from Karel Čapek's 1920 Czech play *R.U.R. (Rossum's Universal Robots)* — coined by Čapek's brother Josef from *robota*, "forced labor" — and that founding story is itself a robot uprising that exterminates humanity. Asimov, writing two decades later, was reacting against this "robot-as-Frankenstein" cliché; the Three Laws were his attempt to imagine machines as engineered products with built-in safety interlocks rather than as doom. He coined **robotics** (first recorded in his 1941 story "Liar!") and **positronic** along the way. The irony of the lineage is that the genre's most famous safety framework was invented to answer the genre's most famous catastrophe — and the catastrophe keeps winning the argument. ## Why it lives in this vault The Three Laws are the headwater of a specific worldview now common among AI founders: that alignment is fundamentally a *rule-writing* problem, and that a sufficiently clever set of constraints (or a sufficiently good objective function) can make a superintelligence safe. Asimov's own fiction is the standing rebuttal to that optimism. The Laws belong here as the canonical case study in [[Science Fiction and the Tech Elite]] — a piece of pulp fiction that became load-bearing in how the people building real AI systems talk about controlling them, often without the cautionary half of the story attached. ## See also - [[Isaac Asimov]] — the author, and his [[Foundation]] saga - [[I, Robot]] — the 1950 collection where the Laws are dramatized - [[Science Fiction and the Tech Elite]] — the hub this page anchors - [[AI Safety]] · [[Alignment]] · [[AGI]] — the modern fields the Laws prefigure - [[Longtermism]] — the present-day descendant of the Zeroth Law's "humanity over the individual" logic