I believe my blog’s subscribers did not receive email notifications of some recent posts. If you’re interested, I’ve listed the links to the last eight posts at the bottom of this edition.

When reading around for my piece yesterday on the wavefunctions of quantum mechanics, I stumbled across an old and fascinating debate about Saturn’s moon Hyperion.

The question of how the smooth, classical world around us emerges from the rules of quantum mechanics has haunted physicists for a century. Most of the time the divide seems easy: quantum laws govern atoms and electrons while planets, chairs, and cats are governed by the laws of Newton and Einstein. Yet there are cases where this distinction is not so easy to draw. One of the most surprising examples comes not from a laboratory experiment but from the cosmos.

In the 1990s, Hyperion became the focus of a deep debate about the nature of classicality, one that quickly snowballed into the so-called Hyperion dispute. It showed how different interpretations of quantum theory could lead to apparently contradictory claims, and how those claims can be settled by making their underlying assumptions clear.

Hyperion is not one of Saturn’s best-known moons but it is among the most unusual. Unlike round bodies such as Titan or Enceladus, Hyperion has an irregular shape, resembling a potato more than a sphere. Its surface is pocked by craters and its interior appears porous, almost like a sponge. But the feature that caught physicists’ attention was its rotation. Hyperion does not spin in a steady, predictable way. Instead, it tumbles chaotically. Its orientation changes in an irregular fashion as it orbits Saturn, influenced by the gravitational pulls of Saturn and Titan, which is a moon larger than Mercury.

In physics, chaos does not mean complete disorder. It means a system is sensitive to its initial conditions. For instance, imagine two weather models that start with almost the same initial data: one says the temperature in your locality at 9:00 am is 20.000º C, the other says it’s 20.001º C. That seems like a meaningless difference. But because the atmosphere is chaotic, this difference can grow rapidly. After a few days, the two models may predict very different outcomes: one may show a sunny afternoon and the other, thunderstorms.

This sensitivity to initial conditions is often called the butterfly effect — it’s the idea that the flap of a butterfly’s wings in Brazil might, through a chain of amplifications, eventually influence the formation of a tornado in Canada.

Hyperion behaves in a similar way. A minuscule difference in its initial spin angle or speed grows exponentially with time, making its future orientation unpredictable beyond a few months. In classical mechanics this is chaos; in quantum mechanics, those tiny initial uncertainties are built in by the uncertainty principle, and chaos amplifies them dramatically. As a result, predicting its orientation more than a few months ahead is impossible, even with precise initial data.

To astronomers, this was a striking case of classical chaos. But to a quantum theorist, it raised a deeper question: how does quantum mechanics describe such a macroscopic, chaotic system?

Why Hyperion interested quantum physicists is rooted in that core feature of quantum theory: the wavefunction. A quantum particle is described by a wavefunction, which encodes the probabilities of finding it in different places or states. A key property of wavefunctions is that they spread over time. A sharply localised particle will gradually smear out, with a nonzero probability of it being found over an expanding region of space.

For microscopic particles such as electrons, this spreading occurs very rapidly. For macroscopic objects, like a chair, an orange or you, the spread is usually negligible. The large mass of everyday objects makes the quantum uncertainty in their motion astronomically small. This is why you don’t have to be worried about your chai mug being in two places at once.

Hyperion is a macroscopic moon, so you might think it falls clearly on the classical side. But this is where chaos changes the picture. In a chaotic system, small uncertainties get amplified exponentially fast. A variable called the Lyapunov exponent measures this sensitivity. If Hyperion begins with an orientation with a minuscule uncertainty, chaos will magnify that uncertainty at an exponential rate. In quantum terms, this means the wavefunction describing Hyperion’s orientation will not spread slowly, as for most macroscopic bodies, but at full tilt.

In 1998, the Polish-American theoretical physicist Wojciech Zurek calculated that within about 20 years, the quantum state of Hyperion should evolve into a superposition of macroscopically distinct orientations. In other words, if you took quantum mechanics seriously, Hyperion would be “pointing this way and that way at once”, just like Schrödinger’s famous cat that is alive and dead at once.

This startling conclusion raised the question: why do we not observe such superpositions in the real Solar System?

Zurek’s answer to this question was decoherence. Say you’re blowing a soap bubble in a dark room. If no light touches it, the bubble is just there, invisible to you. Now shine a torchlight on it. Photons from the bulb will scatter off the bubble and enter your eyes, letting you see its position and color. But here’s the catch: every photon that bounces off the bubble also carries away a little bit of information about it. In quantum terms, the bubble’s wavefunction becomes entangled with all those photons.

If the bubble were treated purely quantum mechanically, you could imagine a strange state where it was simultaneously in many places in the room — a giant superposition. But once trillions of photons have scattered off it, each carrying “which path?” information, the superposition is effectively destroyed. What remains is an apparent mixture of “bubble here” or “bubble there”, and to any observer the bubble looks like a localised classical object. This is decoherence in action: the environment (the sea of photons here) acts like a constant measuring device, preventing large objects from showing quantum weirdness.

For Hyperion, decoherence would be rapid. Interactions with sunlight, Saturn’s magnetospheric particles, and cosmic dust would constantly ‘measure’ Hyperion’s orientation. Any coherent superposition of orientations would be suppressed almost instantly, long before it could ever be observed. Thus, although pure quantum theory predicts Hyperion’s wavefunction would spread into cat-like superpositions, decoherence explains why we only ever see Hyperion in a definite orientation.

Thus Zurek argued that decoherence is essential to understand how the classical world emerges from its quantum substrate. To him, Hyperion provided an astronomical example of how chaotic dynamics could, in principle, generate macroscopic superpositions, and how decoherence ensures these superpositions remain invisible to us.

Not everyone agreed with Zurek’s conclusion, however. In 2005, physicists Nathan Wiebe and Leslie Ballentine revisited the problem. They wanted to know: if we treat Hyperion using the rules of quantum mechanics, do we really need the idea of decoherence to explain why it looks classical? Or would Hyperion look classical even without bringing the environment into the picture?

To answer this, they did something quite concrete. Instead of trying to describe every possible property of Hyperion, they focused on one specific and measurable feature: the part of its spin that pointed along a fixed axis, perpendicular to Hyperion’s orbit. This quantity — essentially the up-and-down component of Hyperion’s tumbling spin — was a natural choice because it can be defined both in classical mechanics and in quantum mechanics. By looking at the same feature in both worlds, they could make a direct comparison.

Wiebe and Ballentine then built a detailed model of Hyperion’s chaotic motion and ran numerical simulations. They asked: if we look at this component of Hyperion’s spin, how does the distribution of outcomes predicted by classical physics compare with the distribution predicted by quantum mechanics?

The result was striking. The two sets of predictions matched extremely well. Even though Hyperion’s quantum state was spreading in complicated ways, the actual probabilities for this chosen feature of its spin lined up with the classical expectations. In other words, for this observable, Hyperion looked just as classical in the quantum description as it did in the classical one.

From this, Wiebe and Ballentine drew a bold conclusion: that Hyperion doesn’t require decoherence to appear classical. The agreement between quantum and classical predictions was already enough. They went further and suggested that this might be true more broadly: perhaps decoherence is not essential to explain why macroscopic bodies, the large objects we see around us, behave classically.

This conclusion went directly against the prevailing view of quantum physics as a whole. By the early 2000s, many physicists believed that decoherence was the central mechanism that bridged the quantum and classical worlds. Zurek and others had spent years showing how environmental interactions suppress the quantum superpositions that would otherwise appear in macroscopic systems. To suggest that decoherence was not essential was to challenge the very foundation of that programme.

The debate quickly gained attention. On one side stood Wiebe and Ballentine, arguing that simple agreement between quantum and classical predictions for certain observables was enough to resolve the issue. On the other stood Zurek and the decoherence community, insisting that the real puzzle was more fundamental: why we never observe interference between large-scale quantum states.

At this time, the Hyperion dispute wasn’t just about a chaotic moon. It was about how we could define ‘classical behavior’ in the first place. For Wiebe and Ballentine, classical meant “quantum predictions match classical ones”. For Zurek et al., classical meant “no detectable superpositions of macroscopically distinct states”. The difference in definitions made the two sides seem to clash.

But then, in 2008, physicist Maximilian Schlosshauer carefully analysed the issue and showed that the two sides were not actually talking about the same problem. The apparent clash arose because Zurek and Wiebe-Ballentine had started from essentially different assumptions.

Specifically, Wiebe and Ballentine had adopted the ensemble interpretation of quantum mechanics. In everyday terms, the ensemble interpretation says, “Don’t take the quantum wavefunction too literally.” That is, it does not describe the “real state” of a single object. Instead, it’s a tool to calculate the probabilities of what we will see if we repeat an experiment many times on many identical systems. It’s like rolling dice. If I say the probability of rolling a 6 is 1/6, that probability does not describe the dice themselves as being in a strange mixture of outcomes. It simply summarises what will happen if I roll a large collection of dice.

Applied to quantum mechanics, the ensemble interpretation works the same way. If an electron is described by a wavefunction that seems to say it is “spread out” over many positions, the ensemble interpretation insists this does not mean the electron is literally smeared across space. Rather, the wavefunction encodes the probabilities for where the electron would be found if we prepared many electrons in the same way and measured them. The apparent superposition is not a weird physical reality, just a statistical recipe.

Wiebe and Ballentine carried this outlook over to Hyperion. When Zurek described Hyperion’s chaotic motion as evolving into a superposition of many distinct orientations, he meant this as a literal statement: without decoherence, the moon’s quantum state really would be in a giant blend of “pointing this way” and “pointing that way”. From his perspective, there was a crisis because no one ever observes moons or chai mugs in such states. Decoherence, he argued, was the missing mechanism that explained why these superpositions never show up.

But under the ensemble interpretation, the situation looks entirely different. For Wiebe and Ballentine, Hyperion’s wavefunction was never a literal “moon in superposition”. It was always just a probability tool, telling us the likelihood of finding Hyperion with one orientation or another if we made a measurement. Their job, then, was simply to check: do these quantum probabilities match the probabilities that classical physics would give us? If they do, then Hyperion behaves classically by definition. There is no puzzle to be solved and no role for decoherence to play.

This explains why Wiebe and Ballentine concentrated on comparing the probability distributions for a single observable, namely the component of Hyperion’s spin along a chosen axis. If the quantum and classical results lined up — as their calculations showed — then from the ensemble point of view Hyperion’s classicality was secured. The apparent superpositions that worried Zurek were never taken as physically real in the first place.

Zurek, on the other hand, was addressing the measurement problem. In standard quantum mechanics, superpositions are physically real. Without decoherence, there is always some observable that could reveal the coherence between different macroscopic orientations. The puzzle is why we never see such observables registering superpositions. Decoherence provided the answer: the environment prevents us from ever detecting those delicate quantum correlations.

In other words, Zurek and Wiebe-Ballentine were tackling different notions of classicality. For Wiebe and Ballentine, classicality meant the match between quantum and classical statistical distributions for certain observables. For Zurek, classicality meant the suppression of interference between macroscopically distinct states.

Once Schlosshauer spotted this difference, the apparent dispute went away. His resolution showed that the clash was less over data than over perspectives. If you adopt the ensemble interpretation, then decoherence indeed seems unnecessary, because you never take the superposition as a real physical state in the first place. If you are interested in solving the measurement problem, then decoherence is crucial, because it explains why macroscopic superpositions never manifest.

The overarching takeaway is that, from the quantum point of view, there is no single definition of what constitutes “classical behaviour”. The Hyperion dispute forced physicists to articulate what they meant by classicality and to recognise the assumptions embedded in different interpretations. Depending on your personal stance, you may emphasise the agreement of statistical distributions or you may emphasise the absence of observable superpositions. Both approaches can be internally consistent — but they  also answer different questions.

For school students that are reading this story, the Hyperion dispute may seem obscure. Why should we care about whether a distant moon’s tumbling motion demands decoherence or not? The reason is that the moon provides a vivid example of a deep issue: how do we reconcile the strange predictions of quantum theory with the ordinary world we see?

In the laboratory, decoherence is an everyday reality. Quantum computers, for example, must be carefully shielded from their environments to prevent decoherence from destroying fragile quantum information. In cosmology, decoherence plays a role in explaining how quantum fluctuations in the early universe influenced the structure of galaxies. Hyperion showed that even an astronomical body can, in principle, highlight the same foundational issues.

---

Last five posts:

1. The guiding light of KD45

2. What on earth is a wavefunction?

3.  The PixxelSpace constellation conundrum

4. The Zomato ad and India’s hustle since 1947

5. A new kind of quantum engine with ultracold atoms

6. Trade rift today, cryogenic tech yesterday

7. What keeps the red queen running?

8. A limit of ‘show, don’t tell’

If you drop a pebble into a pond, ripples spread outward in gentle circles. We all know this sight, and it feels natural to call them waves. Now imagine being told that everything — from an electron to an atom to a speck of dust — can also behave like a wave, even though they are made of matter and not water or air. That is the bold claim of quantum mechanics. The waves in this case are not ripples in a material substance. Instead, they are mathematical entities known as wavefunctions.

At first, this sounds like nothing more than fancy maths. But the wavefunction is central to how the quantum world works. It carries the information that tells us where a particle might be found, what momentum it might have, and how it might interact. In place of neat certainties, the quantum world offers a blur of possibilities. The wavefunction is the map of that blur. The peculiar thing is, experiments show that this ‘blur’ behaves as though it is real. Electrons fired through two slits make interference patterns as though each one went through both slits at once. Molecules too large to see under a microscope can act the same way, spreading out in space like waves until they are detected.

So what exactly is a wavefunction, and how should we think about it? That question has haunted physicists since the early 20th century and it remains unsettled to this day.

In classical life, you can say with confidence, “The cricket ball is here, moving at this speed.” If you can’t measure it, that’s your problem, not nature’s. In quantum mechanics, it is not so simple. Until a measurement is made, a particle does not have a definite position in the classical sense. Instead, the wavefunction stretches out and describes a range of possibilities. If the wavefunction is sharply peaked, the particle is most likely near a particular spot. If it is wide, the particle is spread out. Squaring the wavefunction’s magnitude gives the probability distribution you would see in many repeated experiments.

If this sounds abstract, remember that the predictions are tangible. Interference patterns, tunnelling, superpositions, entanglement — all of these quantum phenomena flow from the properties of the wavefunction. It is the script that the universe seems to follow at its smallest scales.

To make sense of this, many physicists use analogies. Some compare the wavefunction to a musical chord. A chord is not just one note but several at once. When you play it, the sound is rich and full. Similarly, a particle’s wavefunction contains many possible positions (or momenta) simultaneously. Only when you press down with measurement do you “pick out” a single note from the chord.

Others have compared it to a weather forecast. Meteorologists don’t say, “It will rain here at exactly 3:07 pm.” They say, “There’s a 60% chance of showers in this region.” The wavefunction is like nature’s own forecast, except it is more fundamental: it is not our ignorance that makes it probabilistic, but the way the universe itself behaves.

Mathematically, the wavefunction is found by solving the Schrödinger equation, which is a central law of quantum physics. This equation describes how the wavefunction changes in time. It is to quantum mechanics what Newton’s second law (F = ma) is to classical mechanics. But unlike Newton’s law, which predicts a single trajectory, the Schrödinger equation predicts the evolving shape of probabilities. For example, it can show how a sharply localised wavefunction naturally spreads over time, just like a drop of ink disperses in water. The difference is that the spreading is not caused by random mixing but by the fundamental rules of the quantum world.

But does that mean the wavefunction is real, like a water wave you can touch, or is it just a clever mathematical fiction?

There are two broad camps. One camp, sometimes called the instrumentalists, argues the wavefunction is only a tool for making predictions. In this view, nothing actually waves in space. The particle is simply somewhere, and the wavefunction is our best way to calculate the odds of finding it. When we measure, we discover the position, and the wavefunction ‘collapses’ because our information has been updated, not because the world itself has changed.

The other camp, the realists, argues that the wavefunction is as real as any energy field. If the mathematics says a particle is spread out across two slits, then until you measure it, the particle really is spread out, occupying both paths in a superposed state. Measurement then forces the possibilities into a single outcome, but before that moment, the wavefunction’s broad reach isn’t just bookkeeping: it’s physical.

This isn’t an idle philosophical spat. It has consequences for how we interpret famous paradoxes like Schrödinger’s cat — supposedly “alive and dead at once until observed” — and for how we understand the limits of quantum mechanics itself. If the wavefunction is real, then perhaps macroscopic objects like cats, tables or even ourselves can exist in superpositions in the right conditions. If it is not real, then quantum mechanics is only a calculating device, and the world remains classical at larger scales.

The ability of a wavefunction to remain spread out is tied to what physicists call coherence. A coherent state is one where the different parts of the wavefunction stay in step with each other, like musicians in an orchestra keeping perfect time. If even a few instruments go off-beat, the harmony collapses into noise. In the same way, when coherence is lost, the wavefunction’s delicate correlations vanish.

Physicists measure this ‘togetherness’ with a parameter called the coherence length. You can think of it as the distance over which the wavefunction’s rhythm remains intact. A laser pointer offers a good everyday example: its light is coherent, so the waves line up across long distances, allowing a sharp red dot to appear even all the way across a lecture hall. By contrast, the light from a torch is incoherent: the waves quickly fall out of step, producing only a fuzzy glow. In the quantum world, a longer coherence length means the particle’s wavefunction can stay spread out and in tune across a larger stretch of space, making the object more thoroughly delocalised.

However, coherence is fragile. The world outside — the air, the light, the random hustle of molecules — constantly disturbs the system. Each poke causes the system to ‘leak’ information, collapsing the wavefunction’s delicate superposition. This process is called decoherence, and it explains why we don’t see cats or chairs spread out in superpositions in daily life. The environment ‘measures’ them constantly, destroying their quantum fuzziness.

One frontier of modern physics is to see how far coherence can be pushed before decoherence wins. For electrons and atoms, the answer is “very far”. Physicists have found their wavefunctions can stretch across micrometres or more. They have also demonstrated coherence with molecules with thousands of atoms, but keeping them coherent has been much more difficult. For larger solid objects, it’s harder still.

Physicists often talk about expanding a wavefunction. What they mean is deliberately increasing the spatial extent of the quantum state, making the fuzziness spread wider, while still keeping it coherent. Imagine a violin string: if it vibrates softly, the motion is narrow; if it vibrates with larger amplitude, it spreads. In quantum mechanics, expansion is more subtle but the analogy holds: you want the wavefunction to cover more ground not through noise or randomness but through genuine quantum uncertainty.

Another way to picture it is as a drop of ink released into clear water. At first, the drop is tight and dark. Over time, it spreads outward, thinning and covering more space. Expanding a quantum wavefunction is like speeding up this spreading process, but with a twist: the cloud must remain coherent. The ink can’t become blotchy or disturbed by outside currents. Instead, it must preserve its smooth, wave-like character, where all parts of the spread remain correlated.

How can this be done? One way is to relax the trap that’s being used to hold the particle in place. In physics, the trap is described by a potential, which is just a way of talking about how strong the forces are that pull the particle back towards the centre. Imagine a ball sitting in a bowl. The shape of the bowl represents the potential. A deep, steep bowl means strong restoring forces, which prevent the ball from moving around. A shallow bowl means the forces are weaker. That is, if you suddenly make the bowl shallower, the ball is less tightly confined and can explore more space. In the quantum picture, reducing the stiffness of the potential is like flattening the bowl, which allows the wavefunction to swell outward. If you later return the bowl to its steep form, you can catch the now-broader state and measure its properties.

The challenge is to do this fast and cleanly, before decoherence destroys the quantum character. And you must measure in ways that reveal quantum behaviour rather than just classical blur.

This brings us to an experiment reported on August 19 in Physical Review Letters, conducted by researchers at ETH Zürich and their collaborators. It seems the researchers have achieved something unprecedented: they prepared a small silica sphere, only about 100 nm across, in a nearly pure quantum state and then expanded its wavefunction beyond the natural zero-point limit. This means they coherently stretched the particle’s quantum fuzziness farther than the smallest quantum wiggle that nature usually allows, while still keeping the state coherent.

To appreciate why this matters, let’s consider the numbers. The zero-point motion of their nanoparticle — the smallest possible movement even at absolute zero — is about 17 picometres (one picometre is a trillionth of a meter). Before expansion, the coherence length was about 21 pm. After the expansion protocol, it reached roughly 73 pm, more than tripling the initial reach and surpassing the ground-state value. For something as massive as a nanoparticle, this is a big step.

The team began by levitating a silica nanoparticle in an optical tweezer, created by a tightly focused laser beam. The particle floated in an ultra-high vacuum at a temperature of just 7 K (-266º C). These conditions reduced outside disturbances to almost nothing.

Next, they cooled the particle’s motion close to its ground state using feedback control. By monitoring its position and applying gentle electrical forces through the surrounding electrodes, they damped its jostling until only a fraction of a quantum of motion remained. At this point, the particle was quiet enough for quantum effects to dominate.

The core step was the two-pulse expansion protocol. First, the researchers switched off the cooling and briefly lowered the trap’s stiffness by reducing the laser power. This allowed the wavefunction to spread. Then, after a carefully timed delay, they applied a second softening pulse. This sequence cancelled out unwanted drifts caused by stray forces while letting the wavefunction expand even further.

Finally, they restored the trap to full strength and measured the particle’s motion by studying how they scattered light. Repeating this process hundreds of times gave them a statistical view of the expanded state.

The results showed that the nanoparticle’s wavefunction expanded far beyond its zero-point motion while still remaining coherent. The coherence length grew more than threefold, reaching 73 ± 34 pm. Per the team, this wasn’t just noisy spread but genuine quantum delocalisation.

More strikingly, the momentum of the nanoparticle had become ‘squeezed’ below its zero-point value. In other words, while uncertainty over the particle’s position increased, that over its momentum decreased, in keeping with Heisenberg’s uncertainty principle. This kind of squeezed state is useful because it’s especially sensitive to feeble external forces.

The data matched theoretical models that considered photon recoil to be the main source of decoherence. Each scattered photon gave the nanoparticle a small kick, and this set a fundamental limit. The experiment confirmed that photon recoil was indeed the bottleneck, not hidden technical noise. The researchers have suggested using dark traps in future — trapping methods that use less light, such as radio-frequency fields — to reduce this recoil. With such tools, the coherence lengths can potentially be expanded to scales comparable to the particle’s size. Imagine a nanoparticle existing in a state that spans its own diameter. That would be a true macroscopic quantum object.

This new study pushes quantum mechanics into a new regime. Thus far, large, solid objects like nanoparticles could be cooled and controlled, but their coherence lengths stayed pinned near the zero-point level. Here, the researchers were able to deliberately increase the coherence length beyond that limit, and in doing so showed that quantum fuzziness can be engineered, not just preserved.

The implications are broad. On the practical side, delocalised nanoparticles could become extremely sensitive force sensors, able to detect faint electric or gravitational forces. On the fundamental side, the ability to hold large objects in coherent, expanded states is a step towards probing whether gravity itself has quantum features. Several theoretical proposals suggest that if two massive objects in superposition can become entangled through their mutual gravity, it would prove gravity must be quantum. To reach that stage, experiments must first learn to create and control delocalised states like this one.

The possibilities for sensing in particular are exciting. Imagine a nanoparticle prepared in a squeezed, delocalised state being used to detect the tug of an unseen mass nearby or to measure an electric field too weak for ordinary instruments. Some physicists have speculated that such systems could help search for exotic particles such as certain dark matter candidates, which might nudge the nanoparticle ever so slightly. The extreme sensitivity arises because a delocalised quantum object is like a feather balanced on a pin: the tiniest push shifts it in measurable ways.

There are also parallels with past breakthroughs. The Laser Interferometer Gravitational-wave Observatories, which detect gravitational waves, rely on manipulating quantum noise in light to reach unprecedented sensitivity. The ETH Zürich experiment has extended the same philosophy into the mechanical world of nanoparticles. Both cases show that pushing deeper into quantum control could yield technologies that were once unimaginable.

But beyond the technologies also lies a more interesting philosophical edge. The experiment strengthens the case that the wavefunction behaves like something real. If it were only an abstract formula, could we stretch it, squeeze it, and measure the changes in line with theory? The fact that researchers can engineer the wavefunction of a many-atom object and watch it respond like a physical entity tilts the balance towards reality. At the least, it shows that the wavefunction is not just a mathematical ghost. It’s a structure that researchers can shape with lasers and measure with detectors.

There are also of course the broader human questions. If nature at its core is described not by certainties but by probabilities, then philosophers must rethink determinism, the idea that everything is fixed in advance. Our everyday world looks predictable only because decoherence hides the fuzziness. But under carefully controlled conditions, that fuzziness comes back into view. Experiments like this remind us that the universe is stranger, and more flexible, than classical common sense would suggest.

The experiment also reminds us that the line between the quantum and classical worlds is not a brick wall but a veil — thin, fragile, and possibly removable in the right conditions. And each time we lift it a little further, we don’t just see strange behaviour: we also glimpse sensors more sensitive than ever, tests of gravity’s quantum nature, and perhaps someday, direct encounters with macroscopic superpositions that will force us to rewrite what we mean by reality.