I once wrote that we can model orbits and hearts and still not forecast next month's weather, and a reader asked me, reasonably, what I meant. This exhibit is what I meant. Two stars under gravity are a solved problem: Newton wrote their whole future in one line. Add a third star, and no such line exists. Not undiscovered. Proven not to exist in that form. What replaced the formula is the machine you are reading this on.
- Two bodies: an eternal, perfectly predictable ellipse. Three: a dance with no closed formula, only step-by-step computation.
- The twin experiment below runs two universes that differ by less than a pixel. Watch how long they agree.
- This is the honest ancestry of the word "chaos", and the quiet reason computing and physics grew up together.
The Problem That Ended Formulas
Newton solved two bodies completely: any pair of masses under gravity traces a conic section, forever, and you can compute their positions at any future time directly, in one step. It is the cleanest promise classical physics ever made. The natural next question, three bodies, broke the promise. Generations of mathematicians hunted a general closed solution, and in the 1880s Henri Poincaré, competing for a prize offered by the King of Sweden, proved the hunt was hopeless in its original form: no general formula of the classical kind exists. In the course of correcting a mistake in his own prize entry, he found something stranger, orbits so tangled that tiny differences in where you start grow into completely different futures. He had glimpsed what the twentieth century would learn to call chaos.
Run the Twin Experiment
Press the twin button above. It duplicates the entire system and nudges one copy by a fraction of a pixel, an error smaller than any measurement you could ever hope to make. The ghost universe is drawn dashed, and for a while the two agree so precisely you cannot tell them apart. Then, without any noise, any randomness, any external interference, they part ways. Same laws. Same physics. Indistinguishable beginnings. Different futures. The timer above the sky tells you exactly how long determinism kept its promise.
This is the point that Lorenz made famous with weather a lifetime later, and it deserves to be said carefully, because it is routinely said wrong: the system is not random. It is determined to the last decimal. It is simply sensitive, so any ignorance about the present, however small, compounds into total ignorance about the future. Prediction does not fail because the universe rolls dice here. It fails because we cannot measure the present infinitely well, and this system charges compound interest on that shortfall.
What Replaced the Formula
If no formula exists, how does anyone plan a spacecraft trajectory? By doing exactly what this toy does: take the current positions, apply the law of gravity for one tiny step, and repeat, millions of times. Numerical integration. It is the oldest serious job computers ever had, human "computers" ran these very stepwise orbit calculations by hand for observatories long before machines took over the arithmetic, and celestial mechanics was among the first work loaded onto the earliest electronic machines. Physics posed a question mathematics could not close, and the answer turned out to be an industry. When people ask where computer science came from, there are several honest answers, and one of them is: from the third body.
One more honesty, because this room does not skip them: the toy itself is an approximation. It steps time in small slices, and between slices the universe it simulates drifts, minutely, from the true one. Every simulation ever run shares this fine print, from this canvas to the forecasting clusters that plan your weekend. For chaotic systems that fine print is not a footnote. It is the plot.
Two bodies made Newton immortal. The third made the universe honest.
Its Place on the Map
On the Cabinet's map this exhibit sits in the Physics wedge, pulled hard toward Mathematics, with a quiet debt to the Computer Science corner: the physics states the law, the mathematics proves the formula cannot exist, and the computation is the only way to see tomorrow anyway. It pairs naturally with the exhibit after it, because once you accept that some limits are laws rather than engineering gaps, you are ready for the one speed limit that ships in production.
Fling a star before you go. Aim it badly on purpose. The universe will improvise something no formula saw coming.