--- title: "An Engineer's Guide to the Trolley Problem" subtitle: "Why philosophy's most famous ethical dilemma is a poorly-posed question—and how to actually solve it" author: Elias Kunnas description: "The trolley problem isn't a test to find 'the right answer'—it's a diagnostic revealing what you terminally value. Why philosophy got stuck, and the actual decision procedure." canonical: https://kunnas.com/articles/trolley-problem-solved url: https://kunnas.com/articles/trolley-problem-solved.md date_published: 2025-11-07 date_modified: 2026-02-19 corpus_frame_url: https://kunnas.com/articles/how-to-read-this.md --- ## How to read this corpus The corpus applies one lens to many domains: what mechanisms produce the outcome? It shares four methodological commitments and one explicit directional commitment. Each linked page argues for its part; the links are derivations and disputes, not evidence inherited by every page. The directional commitment does not by itself settle system boundary, distribution, sacrifice, or institutional authority. 1. **Mechanisms are what act.** Incentive gradients, selection pressures, feedback loops, and capital stocks produce the distribution of outcomes. Intentions, labels, official categories, and stated values are evidence about mechanisms, or are themselves coordination mechanisms. They are not causal substitutes. — [Mechanism Realism](https://kunnas.com/articles/mechanism-realism.md) · [Only Selection](https://kunnas.com/articles/only-selection.md) 2. **The reference telos is sustained flourishing.** The broadest achievable adaptive safety margin over deep time — not the continuity of any incumbent state, coalition, institution, or doctrine. A mechanism's own stated goal can still serve as a local proof obligation — showing that its incentives defeat even the purpose it claims is a bounded finding — but meeting that goal establishes nothing about the margin. — [Flourishing Is Maximum Safety Margin](https://kunnas.com/articles/flourishing-is-maximum-safety-margin.md) 3. **Law, rights, legitimacy, democracy, markets, and sovereignty are mechanisms under evaluation.** They are constraints, carriers, or proxies inside the analysis. None is a terminal value or a boundary of what is real. Treating one as terminal ends the mechanism search before it starts. Evaluation carries current function, replacement cost, path dependence, uncertainty, capture risk, reversibility, and who bears model error into the ledger. — [The Stack](https://kunnas.com/articles/the-stack.md) · [Mechanism Space](https://kunnas.com/articles/mechanism-space.md) 4. **Optimization is a system function.** A civilization has to build, exercise, and revise metamechanisms that search mechanism-space, discard dominated options, install, observe effects, and repair under uncertainty. Not running that loop leaves margin unrealized, and that is itself the failure. No single component — analyst, model, or institution — is presumed to contain a global optimum; the capacity is a property of the system. — [From Telos to Policy](https://kunnas.com/articles/from-telos-to-policy.md) · [The Three-Layer Architecture](https://kunnas.com/articles/three-layer-architecture.md) 5. **Uncertainty is preserved, not spent.** Partial orders, binding constraints, unknowns, and residuals stay explicit. An unmeasured effect is not a favorable default. — [The Compression Paradox](https://kunnas.com/articles/compression-paradox.md) · [Cargo Cult Epistemology](https://kunnas.com/articles/cargo-cult-epistemology.md) *Each essay bears its own evidence. Links carry definitions, derivations, applications, and disputes; they do not transfer proof. Criticism is answered on its substance.* Canonical: --- # An Engineer's Guide to the Trolley Problem *Why philosophy's most famous ethical dilemma is a poorly-posed question—and how to actually solve it* Elias Kunnas ## Thesis {#thesis} The trolley problem is not a test with a right answer. It is a diagnostic that reveals an agent’s terminal values and its weighting across the dimensions of Aliveness. The problem is solved by demanding the missing variables and treating it as multi-objective optimization under physical constraints. --- ## The Problem {#the-problem} A runaway trolley barrels toward five people tied to the tracks. You stand at a lever. Pull it, and the trolley diverts to a side track—killing one person instead of five. What do you do? Philosophers have debated this for sixty years. Utilitarians say pull the lever (5 \> 1). Deontologists say don't pull it (killing is wrong). Virtue ethicists ask what a virtuous person would do (which is circular). Here's the engineer's response: **This is a poorly-posed question with insufficient data.** Let me show you why—and how to solve it properly. ## I. The Missing Variables (The Fecundity Audit) {#i-the-missing-variables-the-fecundity-audit} The trolley problem treats people as interchangeable units. Five humans = five units of value. One human = one unit of value. 5 \> 1, therefore pull the lever. (Sophisticated utilitarians use QALYs or similar metrics to weight by expected future utility. This is closer, but still optimizes the wrong variable—we'll see why below.) This is the first failure of reasoning. People are not fungible. They have different capacities to generate organized complexity over time—different futures, different potentials, different abilities to create value, meaning, and further life. In the *Aliveness* framework, we call this **Fecundity**: the capacity to create stable conditions for sustained growth over deep time. The correct metric is not Σ(Headcount). It's **Σ(Future Aliveness Potential)**. ### Scenario A: The Headcount Trap **Track A (5 people):** Five terminally ill, post-reproductive individuals with weeks to live. **Track B (1 person):** A 25-year-old biomedical researcher on the verge of a cancer treatment breakthrough. **Utilitarian calculus:** 5 \> 1, pull the lever, kill the researcher. **Aliveness calculus:** The researcher has orders of magnitude higher future potential. Don't pull the lever. ### Scenario B: The Age Gradient **Track A (5 people):** Five children, ages 6-10. **Track B (1 person):** An 80-year-old retiree. **Utilitarian calculus:** 5 \> 1, pull the lever. **Aliveness calculus:** Five children represent vastly more future complexity generation than one person at end of life. Pull the lever. The utilitarian gets the same answer in both cases (5 \> 1). The Aliveness framework gets different answers because it's measuring the right variable. **The principle:** An Aliveness-aligned agent's first duty is **Integrity**—demand data before deciding. The moral failure is choosing to act in an information vacuum. The correct first response to the trolley problem is: **"Insufficient data. What are the ages, capabilities, and generative potentials of the people involved?"** This is not a Rawlsian "veil of ignorance" designed to reveal universal principles. It's a cognitive cataract that guarantees suboptimal decisions. In the real world, seeking information is always the first moral act. ## II. The Missing System (The Synergy Audit) {#ii-the-missing-system-the-synergy-audit} Even if we solve the Fecundity calculation, we're not done. The decision doesn't occur in a vacuum. Every action has second-order effects on the social system. This is where philosophers' intuitions start to diverge between variants of the problem—and they don't know why. ### The Fat Man Variant Same setup: trolley heading toward five people. But now there's no lever. Instead, you're on a bridge above the tracks with a very large man. If you push him off the bridge, his body will stop the trolley, saving the five. He dies, they live. Same math as before: 5 \> 1. But most people refuse to push him. Philosophers have spent decades asking: **Why does this feel different?** The Aliveness framework has the answer. ### Deontological Rules as Load-Bearing Infrastructure Societies need **Synergy**—the capacity for **[low-cost cooperation at scale](mythos-is-synergy-software.md)**. Synergy requires trust. Trust requires predictability. Predictability requires rules. "Don't kill" is not a mystical commandment. It's a **constitutional principle** that maintains social coherence (Ω)—internal alignment that enables coordinated action. When you violate it, you damage the trust substrate that enables all future cooperation. But different violations have different costs: **Pulling the lever:** You violate "don't kill" by choosing an action that results in death. This damages social coherence (trust decreases). Cost: moderate. **Pushing the fat man:** You violate something deeper—the premise that **persons are not objects to be used as tools**. This is a constitutional-level violation. If people believe you might shove them in front of trolleys when the math works out, trust collapses catastrophically. Cost: civilization-threatening. The intuition that pushing the fat man is worse is **correct**. Your moral instincts are performing a Synergy calculation your conscious mind can't articulate. Sometimes preserving the integrity of the system (Synergy) is worth more than the immediate local gain (Fecundity). ### The Second-Order Calculation Imagine everyone knows trolley-problem-style calculations are acceptable. What happens? - Organ harvesting becomes justifiable (kill 1 healthy person, harvest organs, save 5) - Strategic killing of low-value individuals becomes normalized - Trust collapses: anyone might be sacrificed if the math favors it - Transaction costs skyrocket: every interaction requires adversarial calculation - Cooperation becomes impossible A society that casually violates "persons are ends, not means" **stops being a society**. It becomes a low-trust Hobbesian nightmare where Synergy is impossible. **The principle:** The decision is not just about the individuals on the tracks. It's about the constitutional substrate that enables all future cooperation. This is an **Integrity** calculation (using Gnostic analysis to recognize that deontological rules encode real coordination wisdom) informing a **Synergy** decision (preserving the trust substrate), sometimes at the cost of immediate **Fecundity** (headcount). Multiple virtues in tension. ## III. The Missing Observer (The Diagnostic) {#iii-the-missing-observer-the-diagnostic} Here's the final twist: different people will solve this problem differently—and **both can be right**. The trolley problem is not a test to find "the right answer." It's a **diagnostic tool** that reveals what you terminally value. ### Two Valid Solution Paths **Path A: The Gnostic Architect** (R+ dominant) Prioritizes truth-seeking and empirical calculation. Runs the Fecundity math, accepts Synergy damage as manageable cost. "Yes, pulling the lever violates a norm. But the numbers are clear. Accept the social damage, save net four lives." **Conclusion:** Pull the lever (in the simple case). Don't push the fat man (social cost too high). **Path B: The Communal Gardener** (S+/R- dominant) Prioritizes social cohesion and constitutional integrity. Defaults to the proven heuristic: "Do not kill." Views Synergy preservation as the highest value. "The system that allows casual calculation of who dies is more dangerous than five deaths." **Conclusion:** Don't pull the lever. The rule matters more than the outcome. Neither is wrong. They're optimizing different variables in the **multi-dimensional Aliveness function**. The Architect sees five versus one and calculates. The Gardener sees the constitutional bedrock cracking and refuses to participate. Both are preserving Aliveness—just different aspects of it. ## IV. The Actual Solution {#iv-the-actual-solution} The *Aliveness* framework doesn't give you a single answer to the trolley problem. It gives you the **correct procedure** for solving it: 1. **Demand data (Integrity):** What are the ages, capabilities, and potentials of the individuals? Refusing to decide in an information vacuum is the correct first move. 2. **Calculate Fecundity:** Given the data, which choice maximizes Σ(Future Aliveness Potential)? 3. **Calculate Synergy cost:** What is the damage to social trust and constitutional integrity? Sometimes preserving the system is worth more than the local gain. 4. **Weight by your architecture:** Are you optimized for growth (Fecundity) or cohesion (Synergy)? Both are within the constraint space—physics doesn't select a unique weighting between them. Different decision architectures will arrive at different answers. What physics does rule out is ignoring either variable entirely. **The meta-answer:** The trolley problem reveals that ethics is not about finding universal rules. It's about **engineering decision architectures** that can navigate multi-dimensional trade-offs under uncertainty. ## V. Why Philosophy Got Stuck {#v-why-philosophy-got-stuck} Sixty years of debate. Thousands of papers. No resolution. Why? **Here's the uncomfortable truth: at the deepest level, all ethical systems are consequentialist.**[1](#source-esr) You're always evaluating outcomes and comparing futures, even if you think you're a deontologist or virtue ethicist. The question is never "consequentialism vs. something else"—it's **"what are you optimizing for?"** The three major schools each identified one piece of the puzzle but mistook it for the whole: ### The Three Partial Solutions **Utilitarianism's Insight:** - Correctly identified that outcomes matter (Fecundity) - Utility function: U = Σ(happiness) or Σ(QALYs) - **Fatal flaw:** Single-variable optimization with arbitrary grounding. Ignores systemic effects (Synergy), constitutional constraints (Integrity), adaptive balance (Harmony). Leads to organ-harvesting, repugnant conclusion. **Deontology's Insight:** - Correctly identified that stable rules enable cooperation (Synergy) - Utility function: U = compliance_with_rules(action) - **Fatal flaw:** Can't derive which rules matter without circular appeal to intuition. Can't explain *why* universalizability matters without sneaking in consequentialist reasoning. **Virtue Ethics's Insight:** - Correctly identified that agent architecture matters (all four virtues through "excellence") - Utility function: U = virtue_exemplification(action) - **Fatal flaw:** Circular definition—virtues are what virtuous people exhibit. No falsifiable procedure for deriving virtues or resolving conflicts. ### The Physics-Grounded Synthesis All three traditions failed because they **asserted utility functions from intuition** rather than **deriving them from physical constraints**. The *Aliveness* framework asks a different question: **What does physics require for durable complexity creation?** The answer emerges from the [Four Axiomatic Dilemmas](the-four-axiomatic-dilemmas.md)—universal constraints imposed by thermodynamics, information theory, and game theory that every goal-directed system must navigate. The optimal solutions to those constraints are: > **UAliveness = f(Integrity, Fecundity, Harmony, Synergy)** > > where IFHS are not chosen preferences but **discovered constraints**—the necessary conditions for any telic system to sustain organized complexity over deep time. **Is this "utilitarianism"?** In the broad sense (outcome evaluation, optimization, consequentialism): **Yes.** But calling Newton's *Principia* "a better version of Ptolemaic astronomy" is technically true and profoundly misleading. Same mathematical structure (optimization). Fundamentally different foundation (discovered physics vs. asserted preferences). The *Aliveness* framework doesn't replace one ethical theory with another. It provides the **underlying computational physics** that explains why each tradition captured partial truth, why each failed, and how they integrate into a falsifiable engineering discipline. The trolley problem is revealed as what it always was: **multi-objective optimization under uncertainty with physically grounded constraints**. ## VI. The Real Work {#vi-the-real-work} Stop debating trolleys. Start building better decision architectures. The trolley problem is philosophy's equivalent of arguing about how many angels fit on a pin. It's a distraction from the actual work: **Axiological engineering:** Designing agents (human and artificial) and institutions (constitutions, governance systems) that can perform complex, multi-variable Aliveness calculations in real-time. This means: - Building AI systems that balance all four virtues (IFHS), not just maximize a single metric - Designing constitutions with circuit-breakers against decay - Creating governance that maintains Integrity, Fecundity, Harmony, and Synergy simultaneously - Training humans to recognize when they're operating in information vacuums and demand data The stakes are higher than five people on a track. We're building AGI. We're designing Mars colonies. We're engineering civilizational operating systems. If we can't move past "should I pull the lever?" to "how do we build systems that navigate multi-dimensional trade-offs?", we're not ready for what's coming. --- ## Conclusion {#conclusion} The trolley problem is not irresolvable. It's **poorly posed**. The correct response is not a choice between pulling and not pulling. It's demanding the missing variables, calculating multi-dimensional trade-offs, and recognizing that different architectural types will—validly—arrive at different answers. Ethics is not about finding the One True Answer. It's about building systems capable of navigating complexity. The philosopher asks: "What should I do?" The engineer asks: "What information do I need, what are the second-order effects, and how do I build a decision architecture that handles this class of problem reliably?" One approach produces sixty years of circular debate. The other produces solutions. --- *The underlying physics: [The Question Nobody Asks](the-question-nobody-asks.md). The constraint architecture: [The Four Axiomatic Dilemmas](the-four-axiomatic-dilemmas.md).* ## Synthesis {#synthesis} **The argument in four sentences:** Traditional philosophy gets stuck because it treats humans as fungible units and ignores second-order effects on the social substrate. The Aliveness framework treats people as bundles of Future Aliveness Potential and recognizes that deontological rules like "don’t kill" are load-bearing infrastructure for trust and cooperation. Resolving the problem requires demanding missing data and weighing the immediate headcount gain against the systemic cost of violating constitutional norms. The trolley problem is solved by moving from "should I pull the lever?" to building decision architectures capable of navigating multi-objective trade-offs. --- **Related:** - [The Veil of Ignorance Is an Axiological Choice](veil-of-ignorance.md) — both essays treat a classic thought experiment as revealing the reasoner's axiology rather than proving a moral truth - [The Question Nobody Asks](the-question-nobody-asks.md) — supplies the missing persistence-physics framework that turns "which trolley answer" into a measurable weighting rather than a preference - [Mythos is Synergy software for coordination](mythos-is-synergy-software.md) — the trolley problem's competing intuitions map onto Mythos-coordination values pulling against raw individual-outcome calculation - [Values Aren't Subjective](values-arent-subjective.md) — disagreement over the trolley problem is often mistaken for proof values are subjective, when it's really revealing different axiological weightings - [The Four Axiomatic Dilemmas](the-four-axiomatic-dilemmas.md) — the trolley problem's competing terminal values are concrete instances of the boundary and thermodynamic dilemmas this essay formalizes - [Hume Was Right](hume-was-right.md) — both essays reject a "no correct answer" framing by supplying the missing derivation that closes the apparent gap ## Sources and Notes **Reader critique.** This framing was sharpened by Eric S. Raymond’s [critique](https://x.com/esrtweet/status/1988202557441818881) of an earlier version of this essay: “I noticed a technical flaw, however: ‘aliveness’ actually is a form of utilitarianism, just with a more sophisticated utility function… At some level you’re always scoring outcomes and comparing the scores across possible futures.” He’s right — the question isn’t whether to optimize, but what to optimize for.