Summary of "Why Greatness Cannot Be Planned: The Myth of the Objective"

Core Idea

The book’s central claim is that ambitious objectives often mislead more than they guide, because the stepping stones to great outcomes usually do not resemble the final outcome.
The authors argue for a third way between rigid goal-chasing and aimless wandering: intelligent exploration without a fixed objective.
In their view, great discoveries are often found by collecting stepping stones and staying open to what emerges, rather than by following a precise destination.

Why Objectives Fail

The authors call this failure the false compass problem: an objective measures resemblance to the target, not real progress toward the kind of discoveries that make the target possible.
This is especially dangerous in deceptive search spaces, where the best immediate move seems wrong because it temporarily moves away from the objective.
The Chinese finger trap captures the point: the way out is to push inward, just as some problems require moving away from the goal before reaching it.
The maze experiments dramatize this: objective-based search got trapped near the goal, while novelty search succeeded far more often.
In one comparison, novelty search succeeded 39/40 times versus 3/40 for objective-driven search.
Novelty search rewards behaviors for being new relative to past behaviors, not for getting closer to a predefined target.
Because novelty is time-relative, search tends to move from simple to complex as easy novelties are exhausted.
The authors treat this as an information accumulator: a non-objective search gradually encodes facts about the world in the discoveries it preserves.

Examples of Unplanned Greatness

The book repeatedly uses historical examples to show that crucial prerequisites are often invented for other purposes before becoming stepping stones to something bigger.
Vacuum tubes helped lead to computers, but were not invented for computing.
Engines, magnetrons, and many other precursor technologies likewise emerged without their later uses in mind.
Rock and roll is presented as an unplanned cultural emergence from jazz, blues, gospel, and country, rather than something anyone set out to invent.
Elvis Presley’s style is treated as a case of productive accident and serendipity, not a disciplined march toward a preset goal.
The authors extend the same logic to love, arguing that searching too deliberately can undermine what is often found by openness instead.
They defend hobbies and play as valuable stepping stones, using examples like Lego art and Rube Goldberg machines.
Picbreeder, the authors’ interactive image-evolution system, is the clearest model: users often found striking images only when they were not the object of deliberate pursuit.
In the car example, Ken did not aim to evolve a car; he branched from an alien face, and random mutations gradually turned “eyes” into wheels.
The Skull image also emerged through odd stepping stones like a crescent, donut, and dish, showing why a direct “skull-ness” metric would have missed it.

Institutions, Metrics, and Search

The authors argue that objective obsession distorts not just algorithms but institutions, especially when quantitative indicators become targets.
They invoke Campbell’s law: the more a social indicator is used for decision-making, the more it will be corrupted and the more it will distort the process it measures.
Education is the clearest case: standardized testing and accountability encourage teaching to the test, memorization, and metric gaming rather than deep learning.
The same logic is applied to GDP fetishism: maximizing the metric can diverge from improving real social conditions.
The authors stress that even better measurement does not solve the deeper problem if the metric is the wrong compass.
Their critique of Common Core is that a single uniform standard can suppress diversity and experimentation, making future improvement less likely.
Instead of test-driven control, they favor peer review for teachers, where educators assess portfolios, methods, and student work, with standardized testing only as a fallback.
In science, they warn that grant and publication systems can reward consensus over exploration, filtering out interesting but divisive ideas.
The authors cite cases where radical ideas were initially dismissed, such as Darwin and later Kuhn-style paradigm shifts, to show that disagreement can be a sign of importance.
They criticize large directed programs such as the Fifth Generation Computer Systems project and the War on Cancer as examples of ambitious objectives that did not deliver their stated goals, even if they produced byproducts.
They are skeptical of funding only what has immediate impact, because truly useful stepping stones are often not obviously transformative in advance.
Abstract work can matter profoundly later: they mention group theory and public-key cryptography as examples of “useless” ideas that became foundational.

AI, Science, and What to Reward

AI is the chapter’s main scientific case study because it is an explicit search for something highly ambitious, namely intelligence.
The field’s dominant filters are the experimentalist heuristic and the theoretical heuristic.
The experimentalist heuristic favors algorithms that win on benchmarks, but the authors argue this can prune new lines of exploration.
A slightly worse method, like Weird versus OldReliable, may be more valuable because it opens the path to Weirder and future discoveries.
The theoretical heuristic similarly favors algorithms with proofs and guarantees, but a proof for an existing method says little about what may lie beyond its assumptions.
The authors argue that insisting on guarantees can make the field blind to fruitful ideas that do not fit current proof patterns.
Their broader point is that a good algorithm is not just one that performs well now, but one that becomes a stepping stone to better future algorithms.
They use a hypothetical Journal of AI Discovery to show that if reviewers were forced to ignore benchmarks and guarantees, the field would have to confront the limits of its own objective criteria.
As alternatives, they propose judging ideas by inspiration, elegance, creativity, novelty, challenge to the status quo, analogy to nature, beauty, simplicity, and imagination.
Their final stance is not anti-rigor but anti-false-compass: science and AI should rely more on open-minded expert judgment and less on rigid objectives.

What To Take Away

Greatness is often discovered, not targeted: the path to major outcomes usually runs through surprising, indirect stepping stones.
Metrics are dangerous when treated as destinations: they can become false compasses that reward gaming or narrow optimization.
Novelty and diversity matter because they preserve room for discoveries that objective-based systems would prematurely reject.
The book’s deepest recommendation is to treat life, education, science, and innovation as treasure hunts for stepping stones, not as marches toward a single predefined finish line.