Seeing Early, Thinking Sideways

The first two parts of this series were about the weather — the crisis season we're in, and why ideas move through us all at once. This one is about what to do standing in that weather, if what you want to do is build something. It is the most practical thing I have written, and it is aimed squarely at anyone with the itch to start — the founder who hasn't started yet, and the one already in the thick of it.

Here is the whole problem in a sentence. In part two we saw that ideas become ripe — reachable — the moment their ingredients exist, and that when they're ripe, several people reach them at once. So a builder needs exactly two edges. You have to see the ripe idea before the crowd does — thinking ahead of time. And you have to reach the idea most of the crowd can't see — thinking out of the box. Foresight and lateral thinking. Everything else — the pitch deck, the hiring, the burn rate — is downstream of these two.

The good news, and the reason this can be a manual rather than a sermon, is that both are trainable. They look like genius from the outside and feel like method from the inside. What follows is the method: the science of how each one works, three thousand years of worked examples, a recipe box you can keep, and — because this matters more than anyone admits — different recipes for different kinds of heads. Take what fits yours. Ignore the rest.

Part I — Thinking ahead of time

Start with a fact about your own brain that took neuroscience a surprisingly long time to say out loud: you are not mainly a memory machine, you are a prediction machine. A group of psychologists led by Martin Seligman made this case under the banner Homo Prospectus — the human as the animal that lives by simulating futures. Your brain is running forecasts constantly, mostly below awareness. Foresight isn't a mystical gift bolted onto a few people. It's the native function. The skill is doing it on purpose, and well.

And we know a lot about what "well" looks like, because Philip Tetlock spent decades measuring it. In tournaments pitting ordinary people against credentialed experts, his "superforecasters" reliably won — and they shared a style, not an IQ. They think in probabilities, not certainties. They break big questions into smaller answerable ones. They are, in Isaiah Berlin's old image, foxes who know many small things rather than hedgehogs who know one big thing — and, above all, they update fast when the facts move. That's the posture. Not "I know what's coming," but "here's my best current bet, and here's what would change it."

Beyond posture, four concrete moves do most of the work, and I've put them, with their histories, in the recipe box further down: first principles (strip a problem to what must be true, then rebuild — the way Aryabhata reasoned, around 500 CE, that the Earth must be spinning while only appearing still); second-order thinking (never stop at the first consequence — ask "and then what?", the discipline that let Shell's scenario planners imagine a 1970s oil shock before it happened and be ready when it did); reading the S-curve (every technology spreads on the same lazy-S, and knowing where you sit on it is most of timing — this is the Perez pattern from part one, and the reason Netflix mailed DVDs while quietly building streaming, and Kodak, holding the first digital-camera patent, did not); and the pre-mortem (imagine it's a year on and you failed, write the autopsy now — Gary Klein's trick for hearing the doubts a hopeful team swallows). Foresight, in the end, is just refusing to be surprised by things that were forecastable — and being honest that plenty isn't.

Part II — Thinking out of the box

Now the other edge, and it has been badly mystified. "Out of the box" is sold as a personality you either have or don't — the wild creative, the visionary. The research says something far more useful and democratic: creativity is mostly recombination. New ideas are old ideas colliding. The biologist Stuart Kauffman called the frontier of newly-reachable combinations the adjacent possible; Arthur Koestler, decades earlier, called the spark of colliding two unrelated frames bisociation; the cognitive scientist Douglas Hofstadter goes further and argues analogy is "the fuel and fire of thinking" itself. The out-of-the-box idea is usually an in-the-box idea from somebody else's box.

That reframes creativity from a gift into a practice with moves you can run on demand: borrow a solution's shape by analogy (Velcro is just burrs, studied; Kepler cracked the planets by imagining them driven like clockwork); recombine deliberately (Gutenberg's press was a wine press plus a coin punch plus paper — three trades, one machine); constrain on purpose, because scarcity forces the mind off the worn path (Dr. Seuss wrote a classic on a fifty-word bet; the Soviet engineer Genrich Altshuller built an entire inventive method, TRIZ, out of turning contradictions into designs); and incubate — load the problem hard, then walk away, because insight fires on a specific brain circuit that the neuroscientists John Kounios and Mark Beeman have actually watched light up, and it prefers the shower to the desk.

There's one more move that isn't about your head at all — it's about where you stand.

The sociologist Ronald Burt studied where good ideas actually come from inside organisations, and found they cluster at the gaps between groups that don't talk to each other — what he called "structural holes." The person bridging two disconnected worlds has an unfair advantage: they can carry an idea that's old news in one world into another where it's revolutionary. Mark Granovetter had shown the cousin of this with "the strength of weak ties" — your distant acquaintances, not your close friends, bring the non-redundant information. So "out-of-the-box" is often less a trait than a position. David Epstein's Range makes the same case for breadth over depth in messy fields: the generalist who has sampled many boxes out-innovates the specialist trapped in one. If you want more original ideas, the cheapest intervention is not to try harder. It's to go stand somewhere else, next to different people.

And here is the part I find genuinely moving: these moves are not new, not Western, and not the property of any golden age. They repeat across every civilisation that ever thought hard, and some of the sharpest examples are among the oldest.

I want to sit with the Indian entries for a moment, carefully, because they get either ignored or wildly overclaimed and both do a disservice. Around 300 BCE, the scholar Piṅgala, trying to enumerate the rhythms of Sanskrit poetry, used a binary system and an array equivalent to what the West would later call Pascal's triangle and the Fibonacci sequence — combinatorial mathematics, roughly two thousand years early, invented to count poetry. Pāṇini, earlier still, described a whole language as a formal system of ordered rules so algorithmic that computer scientists now read him as a precursor to formal grammars. And Brahmagupta, in 628 CE, wrote down the rules for calculating with zero — the audacity of treating nothing as a number you can compute with, an abstraction so powerful the entire modern world runs on it. No jet planes, no fantasy. Just the real thing, which is astonishing enough: the same handful of thinking moves — abstraction, recombination, formal systems — run by different minds, on different continents, across three millennia. The method travels. That's the whole encouragement of this essay. You are not locked out of it by your geography, your degree, or your century.

Here is the toolkit those examples keep drawing from — the reference box, one move per card:

Part III — Different heads need different recipes

Now the caveat that most "how to think" writing skips, and that I care about most: there is no single recipe, because there is no single kind of mind. Telling a naturally divergent, ten-tabs-open explorer to "just be more rigorous and linear" is as useless as telling a careful analyst to "just be more spontaneous." The research on group problem-solving — Scott Page's work on cognitive diversity — shows why: mixed styles solve hard problems better than uniform ones, even uniformly brilliant ones. Variety isn't decoration. It's the engine. Which means the first move is to know your own grain and cut with it, not against it.

Two things follow from this that are worth saying plainly to a founder. First: hire and partner for the styles you lack, not the ones you have. The Maker who finds a Connector, the Explorer who finds an Analyst — those pairs are quietly why some small teams punch so far above their weight. Second: your weakness has a matching move in the box. An over-thinker doesn't need to become spontaneous; they need to incubate and set a hard stop. A shiny-object explorer doesn't need more discipline in the abstract; they need one constraint. Match the tool to the temperament and the friction drops.

Part IV — The recipe, start to finish

So let me put it together as an actual method — ingredients and steps — because a field manual that never lands the plane is just a reading list. This is the loop I try to run when I'm hunting for something worth building. It braids the two edges together: see early, think sideways.

Ingredients: a real problem you keep bumping into; a habit of reading outside your field; a notebook for parts; and the willingness to be wrong quickly.

1. Read the S-curve. Before anything, ask where the puck is going. What just became possible — a new capability, a fallen price, a changed rule — that wasn't possible three years ago? The adjacent possible only just opened; that's where the ripe ideas are. (Seeing ahead.)
2. Go stand somewhere else. Deliberately collect parts from worlds that don't talk to yours — a different industry, an old idea, another country's solution. Be the broker. Most of your raw material should come from outside your lane. (Thinking sideways.)
3. Recombine and analogize. Smash the parts together. Ask: what works over there that no one has carried over here? The first ten combinations will be obvious; keep going. (Sideways.)
4. Constrain hard. Force the idea through a brutal limit — do it for a tenth of the budget, for one narrow user, in a week. Constraint is what turns a vague notion into a specific, testable one. (Sideways.)
5. Run a pre-mortem and go second-order. Now attack it. Imagine it failed — why? And if it works, then what happens two moves later? This is where foresight saves you from a clever idea that dies on contact with reality. (Ahead.)
6. Incubate, then ship a small test. Sleep on it once — genuinely, the off-line mind earns its keep here — then get the smallest real version in front of a real person. An idea you never test is just a nicer opinion.
7. Update like a fox. Treat the result as evidence, not a verdict. Adjust your bet and run the loop again. The founders who win are rarely the ones who were right first; they're the ones who were wrong fastest and cheapest.

Two disciplines keep this honest. Thinking ahead is not fortune-telling — Amara's law from part one still rules, and the fox hedges. And thinking sideways is not novelty for its own sake; a strange idea that meets no real need is just strange. The out-of-the-box idea has to solve an in-the-real-world problem. As Pasteur put it, chance favours the prepared mind — but only the prepared one. The reps are the preparation.

Why this is the democratic part

I'll close where the series keeps returning. The winter we're in rewards exactly these two skills — seeing early and thinking sideways — because that's what a moment of broken assumptions and ripe possibilities is: an open field for anyone who can read it. And the deepest reason for hope is the one the timeline makes plain. These moves are not a modern Western trick or a gift handed to a lucky few. They were run by altar-builders in Vedic India and radar engineers in wartime America and two naturalists an ocean apart. They are ancient, portable, and learnable, which means the edge is not reserved for the pedigreed. A curious person anywhere, with a method and enough reps, can see what the credentialed crowd misses.

You don't need to be a genius. You need a recipe, the humility to know which one suits your head, and the stubbornness to run it more times than the people who quit. That has always been the deal. It's just that in a winter, the payoff for taking it goes up.

(These essays are becoming, slowly, the chapters of something longer — a book on building and thinking in this strange, hopeful decade. If that's a thing you'd want to read, that's useful for me to know.)

Sources & further reading: Martin Seligman, Peter Railton, Roy Baumeister & Chandra Sripada, Homo Prospectus (2016), and Daniel Schacter on episodic future thinking; Philip Tetlock & Dan Gardner, Superforecasting (2015) and Tetlock's Expert Political Judgment (2005), with Isaiah Berlin's fox/hedgehog; Pierre Wack's Shell scenario planning (Harvard Business Review, 1985) and Peter Schwartz, The Art of the Long View (1991); Everett Rogers, Diffusion of Innovations (1962) and Carlota Perez (2002); Gary Klein on the pre-mortem; Stuart Kauffman's adjacent possible via Steven Johnson (2010); Arthur Koestler, The Act of Creation (1964); Dedre Gentner and Douglas Hofstadter on analogy; Edward de Bono, Lateral Thinking (1967) and J.P. Guilford on divergent thinking; Genrich Altshuller's TRIZ; John Kounios & Mark Beeman, The Eureka Factor (2015); Mark Granovetter, "The Strength of Weak Ties" (1973) and Ronald Burt, Structural Holes (1992) and "Structural Holes and Good Ideas" (2004); David Epstein, Range (2019); Scott Page, The Difference (2007); and for the historical leaps, standard histories of the Śulba Sūtras (Baudhāyana), Piṅgala's Chandaḥśāstra, Pāṇini's Aṣṭādhyāyī, Aryabhata's Āryabhaṭīya (499 CE) and Brahmagupta's Brāhmasphuṭasiddhānta (628 CE). Datings are given conservatively; the Vedic-era achievements are remarkable without embellishment, and embellishment is exactly what this essay avoids.