# Why Greatness Cannot Be Planned: The Myth of the Objective *Kenneth O. Stanley & Joel Lehman · 2015 · Springer*  > [!abstract] In one sentence > For any sufficiently ambitious goal, optimizing toward an explicit objective is usually self-defeating, because the stepping stones that actually lead to greatness rarely resemble the goal — so open-ended search for novelty and interestingness discovers more than directed search ever could. ## Thesis Stanley and Lehman make a counterintuitive claim with surprisingly hard backing: pursuing an ambitious objective by measuring and optimizing progress toward it is, for genuinely novel achievements, usually the wrong strategy. The reason is structural, not motivational. The "stepping stones" that lead to a great discovery typically look nothing like the discovery itself — vacuum tubes were a stepping stone to computers, not a worse computer — so any metric that scores candidates by their resemblance to the goal will actively steer the searcher *away* from the path that reaches it. The authors call this the **objective paradox**: the harder and more interesting the goal, the more likely its objective function is deceptive. Their proposed alternative is to abandon the objective and instead search for what is *novel* and *interesting*, collecting stepping stones for their own sake. Greatness, on this view, is something you stumble into from a rich foundation of accumulated diversity — not something you march toward. ## Key Concepts **The deceptive objective.** A fitness function or KPI defines a gradient: "do more of what scores higher." But in most interesting search spaces, the local gradient points into dead ends. Their image is the Chinese finger trap — the direct action (pulling apart) is exactly the one that fails. Objectives lie about which moves are progress. **Stepping stones and the stepping-stone collector.** Innovations are reached through chains of intermediate states whose value is only legible in hindsight. A good search process is a *collector* of these stones, accumulating capabilities and forms without demanding that each one justify itself against a final target. **Novelty search.** The book's empirical anchor, drawn from the authors' AI research. Rather than rewarding agents for getting closer to a goal (e.g., reaching the end of a maze), novelty search rewards them simply for behaving *differently* from everything seen so far. Counterintuitively, novelty search often solves the goal faster than goal-directed search — because it relentlessly expands the frontier of stepping stones instead of getting trapped by deceptive gradients. **Picbreeder.** A collaborative online system where users breed images via the NEAT neuroevolution algorithm, selecting whatever catches their eye. Recognizable images (a butterfly, a car, a face) emerged with no one aiming for them — and crucially, their evolutionary precursors did not resemble the result. This is the book's existence proof that open-ended, interestingness-driven search discovers what objective search cannot. **Treasure hunter vs. objective-driven planner.** The recommended stance is the opportunistic explorer who follows promising leads, contrasted against the planner who locks onto a destination and grinds toward it. ## Intellectual Context The book is unusual: a popular-science argument written by two computational scientists, grounded in concrete algorithms (NEAT, novelty search) rather than anecdote. It generalizes findings from evolutionary computation and open-ended AI into a manifesto about creativity, science funding, education, and careers. It resonates with evolutionary biology's lack of a target, with Hayek-style distributed discovery, and with critiques of metric-driven institutions, while staying anchored in reproducible experiments. ## Reception & Critiques The book is widely cited in the AI/AGI and open-endedness communities and influenced later work on quality-diversity and open-ended learning. Common pushbacks: the thesis is strongest in deceptive, exploratory domains and weakest where objectives are tractable and the gradient is honest (well-posed engineering, near-term goals); "interestingness" is itself hard to define and can smuggle in implicit objectives; and "abandon objectives" is easy to over-apply as an excuse for aimlessness. The authors largely concede the scope limits — their target is *ambitious, deceptive* goals, not all goals. ## On This Shelf This shelf circles a shared suspicion: that the most important order is generated, not imposed. The book is the **algorithmic cousin of Kauffman's *Origins of Order*** — both argue rich structure self-organizes without a designer steering toward it. It rhymes with **Agüera y Arcas's *What Is Intelligence?***, where intelligence emerges from open-ended computation rather than goal specification, and with the open-ended creativity in **Pirsig's *Zen and the Art of Motorcycle Maintenance*** and the elusive "Quality" that resists being reduced to a metric. Against the **Landau-Lifshitz volumes** — the apotheosis of deriving everything from clean variational objectives — it marks the boundary where honest gradients give way to deceptive ones. **MacIntyre's *After Virtue*** supplies the ethical mirror: goods *internal* to a practice, pursued for their own sake, versus external objectives that corrupt it. **Wilber** and **Rosen** share its developmental, anti-reductionist read of how novelty and life unfold; **Zurek's Quantum Darwinism** offers a physics-level analogue of selection without a planner. ## Related Pages - [[origins-of-order-kauffman|The Origins of Order: Self-Organization and Selection in Evolution]] — Kauffman's biological case that complex order self-organizes without a designer is the natural-science counterpart to Stanley & Lehman's algorithmic claim that diversity and structure accumulate without an objective steering them. - [[what-is-intelligence|What Is Intelligence? Lessons from AI About Evolution, Computing, and Minds]] — Both treat intelligence and capability as emergent products of open-ended, non-goal-directed processes rather than something specified by an explicit target. - [[zen-and-the-art-of-motorcycle-maintenance|Zen and the Art of Motorcycle Maintenance: An Inquiry into Values]] — Pirsig's 'Quality' that resists reduction to a metric parallels the book's argument that interestingness and greatness escape capture by an objective function. - [[after-virtue|After Virtue: A Study in Moral Theory]] — MacIntyre's distinction between goods internal to a practice (pursued for their own sake) and external objectives is the moral-philosophy analogue of stepping-stone collecting versus deceptive objective-chasing. - [[landau-lifshitz-mechanics|Mechanics (Course of Theoretical Physics, Volume 1)]] — Landau-Lifshitz exemplifies deriving rich results from clean variational objectives; the book marks where such honest gradients give way to deceptive ones in harder search spaces. - [[statistical-physics-part-1-landau-lifshitz|Statistical Physics, Part 1 (Course of Theoretical Physics, Vol. 5)]] — Statistical physics' account of macroscopic order arising from many interacting parts mirrors novelty search's emergent expansion of the frontier without a guiding target. - [[classical-theory-of-fields-landau-lifshitz|The Classical Theory of Fields (Course of Theoretical Physics, Vol. 2)]] — Another exemplar of objective/principle-driven derivation, contrasting with the book's case for objective-free exploration in deceptive domains. - [[life-itself-rosen|Life Itself: A Comprehensive Inquiry into the Nature, Origin, and Fabrication of Life]] — Rosen's anti-reductionist account of how life resists mechanical specification echoes the book's claim that genuine novelty cannot be planned via a fixed metric. - [[decoherence-and-quantum-darwinism|Decoherence and Quantum Darwinism: From Quantum Foundations to Classical Reality]] — Zurek's selection-without-a-planner — stable classical states 'surviving' decoherence — is a physics-level analogue of novelty/selection processes producing structure without a goal. - [[sex-ecology-spirituality|Sex, Ecology, Spirituality: The Spirit of Evolution]] — Wilber's developmental, evolutionary unfolding of ever-greater complexity shares the book's open-ended, anti-reductionist framing of how novelty emerges over time. --- [[_Index|← Bookshelf Wiki Index]] · [[Synthesis|Cross-cutting Synthesis →]]