I have a complicated relationship with time. There never seems to be enough of it when I’m trying to finish something, yet somehow a microwave can make thirty seconds feel like an entire afternoon. So obviously, when I started reading about scientists reversing quantum processes, my brain immediately went somewhere considerably more ambitious than the experiment.
Could we get a little rewind button? Maybe take back the thing we said before thinking? Rescue a glass of wine before it hits the floor? I’d settle for returning to the moment before I opened another research tab and lost track of what I was supposed to be doing.
The actual science is smaller than that, but still wonderfully strange. Researchers in Vienna demonstrated a way to reverse an unknown change to a qubit encoded in a photon, restoring its earlier quantum state. They used a carefully controlled optical experiment involving a quantum switch. They didn’t send the laboratory into yesterday, stop the universe’s clocks or create a machine you can climb into.
I think that distinction makes the story better once you understand it. We don’t need to add time-traveling astronauts to make a particle of light interesting. Apparently it already has enough going on.
What Does Quantum Time Reversal Actually Mean?
The phrase sounds like a very small time machine, which is probably why it spread so easily. In this context, though, reversal means undoing a particular kind of change in a quantum system’s state. The experiment happens while the equipment, researchers and rest of the room continue moving forward through ordinary time.
Imagine turning a little arrow through a series of rotations and then applying a transformation that restores its original orientation. You’ve reversed what happened to the arrow. You haven’t made the clock on the wall run backward or removed the experience from whoever watched you do it.
A quantum state is considerably more complicated than an arrow on a desk, so keep the comparison loose. The useful part is the distinction between restoring a configuration and revisiting a moment. You can restore your phone’s previous settings this afternoon without making it yesterday again. Quantum rewinding is much stranger than changing phone settings, but those are still two different ideas.
The research asks how to reverse a transformation even when you don’t know exactly which transformation you’re dealing with. That’s where I stopped thinking this was simply fancy wording for an undo button. How do you undo something when you weren’t given the instructions for what happened in the first place?
The Experiment Used a Photon’s Polarization
A photon is a particle of light. Polarization gives researchers a way to encode quantum information in light, using two distinguishable basis states and their superpositions. That two-level system is a qubit, the basic unit we keep hearing about in quantum computing.
A familiar computer bit can be recorded as zero or one. A qubit’s state includes amplitudes associated with its basis states, and the relationship between those amplitudes matters. When scientists manipulate and measure it, the outcomes follow the rules of quantum mechanics.
I find it helpful to remember that a qubit describes a system we’re choosing to use for information. It isn’t a new species of tiny ball that researchers discovered hiding behind an electron. Different physical platforms can provide qubits, including certain properties of photons or carefully controlled atoms and electronic devices.
In this experiment, the team controlled a property of light. They weren’t rewinding every property of a photon’s entire existence, and they weren’t gathering up everything that had ever interacted with it. The scale and the particular system matter enormously.
Sometimes the phrase “scientists reversed a particle” gives me the mental image of a little glowing dot backing down a hallway like a confused delivery driver. That isn’t a useful picture here. The important thing is the quantum state and the transformation applied to it.
Why Reversing Something Unknown Is Such a Clever Problem
If somebody tells you exactly how they changed a system, you may be able to work out how to undo that change. Rotate something one way, then apply the corresponding rotation the other way. Knowing the operation gives you a starting point.
In ideal quantum theory, the evolution of a closed system is described by a unitary operation. A unitary transformation has an inverse. Mathematically, there’s a way back. Actually implementing that inverse when you don’t know the original operation is another question entirely.
This is the part that reminds me of tasting wine blind. Knowing the label and explaining the bottle is one task. Being handed something without the label and asked to do something useful with it is a different task. The bottle hasn’t become magical because I’m missing information, but the information I have changes the problem.
Quantum mechanics also limits the information you can get from a single unknown state. You can’t simply examine it from every possible angle, write down a perfect description and leave it untouched. Measurements give particular outcomes, and the measurement can change the state.
That doesn’t mean consciousness is doing something mystical to reality. A detector is a physical system interacting with another physical system. The point is that measuring quantum information isn’t equivalent to quietly reading a label that was sitting there all along.
So the interesting challenge becomes: can a controlled physical protocol perform the reversal without first learning the complete state and the complete operation? The Vienna work showed a way to do that for the qubit processes they studied.
A Quantum Switch Makes the Order of Operations Interesting
The name quantum switch makes it sound like a slightly expensive light switch. It’s actually about controlling the order in which operations act on a system. The order can itself participate in a quantum superposition.
For a normal sequence, an object experiences operation A and then operation B, or B and then A. In a quantum-switch arrangement, a control system coherently links those different orders to different alternatives. Interference between the alternatives can reveal or use information that a single fixed sequence doesn’t provide.
Order matters because quantum operations don’t generally commute. That’s the technical way of saying doing one thing and then another can give a different result from doing them in the opposite order. It sounds intimidating until you remember how much of everyday life already has this quality.
Put your socks on and then your shoes and you’re ready to leave. Put your shoes on and then your socks and you’ve created a situation people are going to have questions about. Obviously socks aren’t quantum operators, but the importance of order is already familiar.
The quantum part is keeping alternatives coherent so their amplitudes can interfere, rather than simply flipping a coin to decide which order to use. A random choice between two ordinary sequences isn’t the same resource as their coherent quantum control.
That also doesn’t mean a scientist can casually decide that tomorrow causes yesterday. The phrase “order of operations” refers to a specific experimental arrangement. Your appointment still comes after the phone call you made to book it, unfortunately.
What the Vienna Team Actually Demonstrated
The study, Demonstration of universal time-reversal for qubit processes, was published in Optica in 2023. Peter Schiansky and colleagues worked with experimental and theoretical teams including Philip Walther and Miguel Navascués.
The researchers used optical fiber components and free-space interferometers arranged as a quantum switch. Their photonic demonstration restored earlier polarization states without needing to identify the unknown process in the way an ordinary learn-then-undo approach would require.
They reported an average rewinding fidelity above 95%. Fidelity describes how closely the recovered state matches the intended state. It isn’t a claim that 95% of the laboratory went backward in time, or that every possible input will be recovered perfectly on every attempt.
Success probability is another quantity. Their recursive protocol can be extended to make success arbitrarily likely under its theoretical assumptions. That’s different from saying a finite laboratory run has no possibility of failure or no practical cost.
I know those qualifications sound less dramatic than “scientists control time,” but think about what’s still left after you remove the exaggeration. They manipulated a system to reverse a change they hadn’t needed to fully characterize first. That’s a pretty impressive afternoon’s work.
The research team’s explanation is a useful companion to the paper if you’d like to read how the people doing the experiment describe it.
AI-generated conceptual illustration above. The light ribbons are a visual metaphor, not a diagram of the experiment.
“Universal” Has a Smaller Meaning Than It Sounds Like
This word does a lot of work in quantum headlines. Universal makes me think of a remote that operates every television, or a wine that somehow goes with an entire tasting menu. In physics, the relevant question is always: universal across which class of systems or operations?
Here it doesn’t mean there’s a box that reverses absolutely anything placed inside. The protocol concerns a specified quantum setting, with access to operations and controls that meet its assumptions. Those assumptions are part of the result, not a footnote we can throw away once the headline gets exciting.
Related theoretical research in Physical Review Letters describes qubit rewinding using repeatable operations, interference and signals that identify successful outcomes. The word repeatable matters. An operation can be unknown to the person using it while still being available to apply again in a controlled way.
Unknown isn’t the same thing as inaccessible. If someone gives you access to a device, you may be able to use its behavior without possessing its entire internal blueprint. That possibility is much more interesting to me than pretending the equipment has no requirements at all.
There are also different theoretical routes to inversion. A separate 2023 paper describes an exact, deterministic qubit-unitary inversion protocol using four calls to the unknown operation. That doesn’t turn every other experiment into the same method. Different resources produce different possibilities.
I’m including this because several distinct ideas get bundled together online under “quantum time control.” A quantum switch, a quantum time flip, a rewinding protocol and a general statement about time travel aren’t interchangeable labels. You can be impressed by one without accidentally claiming the researchers demonstrated all four.
Can Scientists Pause Time or Speed It Up?
The careful answer depends on the protocol and on what “time” means in the claim. Quantum control can change how a system evolves. A system can be manipulated so its final state corresponds to a different amount or direction of evolution than it would have experienced if simply left alone.
That isn’t permission to say the entire flow of time around the apparatus has been paused. A clock beside it keeps ticking while researchers operate the equipment. You also don’t get to skip the resources, interactions or waiting periods required to implement the protocol.
The Vienna experiment discussed here demonstrated rewinding of qubit processes. I’m keeping the focus there instead of turning a collection of related proposals into a claim that one machine now has play, pause, rewind and fast-forward buttons for reality.
If I put a soup in the freezer and slow down some of the processes that would spoil it, I haven’t paused the afternoon. That’s obviously a very different kind of physics, but it helps me separate changing a process from changing the surrounding world’s history.
The quantum version deserves its own precise explanation. It doesn’t become less remarkable because the researchers had a specific task. I think we sometimes expect science to provide the entire fantasy at once, when a well-defined piece can already be enough to change what we understand.
This Isn’t a General Cure for Quantum Computer Errors
Quantum computers need delicate information to survive long enough for useful operations. The same systems that researchers need to control are also vulnerable to unwanted interactions. Noise can affect preparation, gates, storage and readout. Building reliable devices means managing all of those stages.
Some errors can be described as unwanted coherent rotations. Other noise involves information becoming entangled with an environment we don’t control. Those situations aren’t equivalent. Reversing an unknown unitary operation doesn’t automatically recover information lost through every possible kind of noise.
It’s tempting to imagine a quantum computer noticing something went wrong and casually rewinding the mistake. I would also like my laptop to have this energy whenever it decides to restart during something important. But useful error correction has to preserve the information you want while detecting and addressing the errors.
Quantum error-correcting codes do that by encoding logical information across physical systems and extracting appropriate error information. They don’t simply read out the unknown protected state, write it in a notebook and reconstruct it after a failure.
Rewinding research could contribute useful tools and ideas to quantum information science. The researchers describe potential future applications in that field. That’s a reasonable possibility to discuss. Declaring that this experiment has solved quantum error correction would be a completely different claim.
If you’d like the technical background without the time-machine headlines, IBM’s explanation of quantum error correction shows why the problem requires such careful handling.
Why Not Just Make a Backup Copy of the Qubit?
This was one of the first questions that occurred to me. If the information is so fragile, why not save another copy before doing anything to it? We do that with documents, photographs and all the other things we remember to back up approximately five minutes after losing them.
Quantum information has a restriction called the no-cloning theorem. There’s no operation that perfectly copies every arbitrary unknown quantum state. You can copy certain known or distinguishable states, but that isn’t the same thing as a universal copying machine for unknown qubits.
That’s a fairly rude rule if your entire plan was to treat quantum information like a folder on your desktop. It also explains why the field has so many clever workarounds that don’t resemble the solutions we’re used to using on ordinary computers.
Error correction gets around the problem through encoding, rather than making independent perfect copies of the unknown state. Rewinding explores another set of questions about transformations and the resources required to invert them. Neither gives us permission to ignore the mathematical restriction.
I actually like this part of physics. You can have this wonderfully strange possibility in one direction and a very firm limit right beside it. The universe doesn’t have to organize its rules according to what would be convenient for us.
What About Entangled Particles?
Entanglement is another word that gets invited into almost every quantum story, usually followed by a sentence suggesting two particles are texting each other faster than light. That description is misleading.
Entangled systems can have correlations that can’t be explained by treating each part as having its own independent quantum state. But an experimenter can’t choose a measurement outcome on one side and use it to send a controllable faster-than-light message to the other.
That matters here because “reversing time” and “instant communication” make a very dramatic combination if you put them in the same paragraph. The experiment doesn’t establish a way to message the past, cause distant events at will or let an effect arrive before the decision that produces it.
Caltech’s entanglement explainer is particularly clear about the communication limit. I’ve also written about quantum entanglement separately because it deserves space to be strange on its own terms.
Why Doesn’t This Unscramble an Egg?
The old egg example always shows up in discussions of time, which I suppose is fair. We have extensive personal experience of eggs becoming breakfast and remarkably little experience of breakfast returning to an intact shell.
On everyday scales, an enormous number of microscopic arrangements correspond to something we’d recognize as scrambled eggs. Far fewer correspond to the particular intact egg we started with. Heat, motion and interactions spread information into the surrounding environment while the process unfolds.
The thermodynamic arrow of time is connected to entropy and to the conditions from which systems evolve. Reversing a controlled transformation on a small quantum system doesn’t demonstrate a reversal of the thermodynamic history of an entire kitchen.
It also doesn’t make a refrigerator a violation of the second law. A refrigerator can reduce the temperature of its contents while using energy and releasing heat elsewhere. Looking only at the part that becomes more organized or cooler leaves out the rest of the physical process.
The laboratory has its own equipment, detectors, energy use and records. Recovering one target state doesn’t erase those things. The wider experiment continues to happen, and the researchers don’t have to forget that they performed it.
I think that’s the part my brain keeps trying to skip. We imagine a single object surrounded by blankness, when an actual experiment is a whole arrangement of interacting systems. Choosing what counts as the target is part of understanding what was reversed.
There Isn’t a Countdown to a Person-Sized Version
My earlier version of this article suggested that rewinding a person might require millions of years of processing. That wasn’t a supported estimate, and it gave the wrong impression that we already know a route to doing it if we can just buy enough computers.
We don’t. A living body continually exchanges heat, matter and information with its surroundings. Moving from a controlled qubit process to reversing someone’s biological history isn’t a straightforward scaling exercise with an established completion time.
This experiment didn’t demonstrate reversing aging, recovering a lost memory, revisiting an event or changing a decision after it happened. I’d love to borrow a few hours from a week when I didn’t have so many projects, but that isn’t among the available settings.
There’s room for imagination here without pretending imagination is an experimental result. I can be fascinated by what the research makes me think about while keeping the actual achievement clear. Honestly, the real achievement is already something I would have assumed was impossible before reading about it.
The Small Result Is the Part I Keep Thinking About
What I love about this story is how specific the question became. Someone looked at the problem of undoing an unknown quantum transformation and asked what resources would make it possible. Then other people built an optical setup precise enough to test that idea.
Somewhere between the equations and the equipment, light passed through the experiment and came out in a state that carried evidence of the reversal. A tiny, carefully measured result that tells us something about what nature allows.
I’m apparantly very easy to distract with that sort of thing.
I still don’t get to rescue the wine glass after it shatters. The microwave’s final thirty seconds remain a personal endurance event, and every tab I opened while researching this is still waiting for me. Time has offered no special accommodations.
But I understand a little more about why an earlier state and an earlier moment aren’t the same thing. I also have another reason to be impressed by a particle of light, which is probably a fairly good outcome for an evening spent reading physics I don’t entirely understand on the first try.
The clock keeps going. Somewhere, a photon’s carefully controlled state can be brought back. I think that’s plenty of wonder for one story.
A sand timer on the desk would give me a very tangible way to watch a few minutes go by. Turn it over, give one task my attention and let the physics questions wait beside the keyboard for a moment.
Related Reads You Might Enjoy:
The Invisible Symphony: How the Universe Flickers Through Our Lives Without Us Knowing
Farming the Stars: India’s Space-Grown Superfoods and the Future of Cosmic Agriculture
Where the Universe Hides Her Skeleton: The Tale of Missing Matter
Through the Glass of Light: Microsoft’s Vision of Holographic Teleportation
Necrobotics: The Wild Science of Turning Dead Spiders into Robotic Grippers
When the Internet Lost Its Way: The 18 Minutes That Haunted Cyberspace
The Mouse Utopia That Turned into Hell: What Universe 25 Reveals About Us
Sources and Further Reading
Schiansky and colleagues, Optica (2023)
University of Vienna: Experiment and measured fidelity
Trillo and colleagues: Universal quantum rewinding
Yoshida and colleagues: Exact qubit-unitary inversion
Chiribella and Liu: Quantum operations with indefinite time direction
IBM Quantum: Quantum circuits and no-cloning
IBM Quantum: Correcting quantum errors
