Quantum Tools for Scientific Breakthroughs

Quantum tools for scientific breakthroughs do not require fault-tolerant quantum computers. Squeezed light already improves LIGO's sensitivity, and spin-squeezed atomic clocks already beat the standard quantum limit, while a useful quantum computer for quantum gravity may still be years away.

What quantum tools for scientific breakthroughs already deliver

Quantum tools for scientific breakthroughs are already producing results: squeezed light improves the sensitivity of gravitational-wave detectors, and entangled atomic clocks measure time below the standard quantum limit. A panel recorded at MIT in 2025 covered these sensors alongside neutral-atom quantum computing and quantum error correction used to study black holes.

The panel featured Nergis Mavalvala, the Marble Professor of Astrophysics and Dean of Science at MIT, who works on the Laser Interferometer Gravitational-Wave Observatory (LIGO); Vladan Vuletic, the Lester Wolfe Professor of Physics at MIT and director of the Center for Ultracold Atoms; and Daniel Harlow, an associate professor of physics at MIT who studies quantum gravity. Mavalvala's session described light-based quantum sensing, Vuletic's covered atomic clocks and neutral-atom computing, and Harlow's covered black holes and the emergence of spacetime.

The clearest dividing line on the panel was between sensing and computing. Quantum sensing has shipped. Quantum computing, by the panelists' own account, has not yet reached the size where it answers the questions that motivated it.

Vuletic closed his talk with a third thread that sits apart from both: machine learning used to tune experiments. In his lab, an optimization program beat a month of manual tuning overnight, improving a Bose-Einstein condensate preparation by a factor of 6, a first-hand lab result rather than a published benchmark.

How squeezed light upgraded LIGO

Squeezed light is a quantum state of light engineered so that its uncertainty is smaller in one property than in another, and it raised the sensitivity of Advanced LIGO by 6 dB, a factor of two, during the third observing run. Mavalvala presented that figure and called the gap between the unsqueezed and squeezed sensitivity curves the quantum advantage in the instrument.

LIGO's detectors are L-shaped laser interferometers with 4-kilometer arms. Light travels down both arms, recombines at a beam splitter, and the instrument measures the phase shift caused by mirror displacements from a passing gravitational wave. Vacuum fluctuations entering the interferometer's open port set two limits that pull in opposite directions: shot noise, read out at the photodetector, and radiation pressure noise, where the same fluctuations kick the mirrors.

Mavalvala traced the idea to Carl Caves, who proposed squeezing the input light in the early 1980s. Squeezing turns the round uncertainty region of the vacuum into an ellipse with the same area, so the least noisy axis can be aligned with the signal. Because the ellipse helps at high frequencies and hurts at low ones, LIGO later added a frequency-dependent rotation of the squeezing ellipse, which removed the low-frequency penalty the first version introduced.

Entangled atomic clocks and the standard quantum limit

Entangled atomic clocks use quantum correlations between atoms to reduce measurement noise below what independent atoms can achieve, and Vuletic reported a 12 to 13 dB enhancement, meaning one entangled measurement can replace about 15 unentangled ones. That gain translates directly into a faster data acquisition rate.

The clock design holds laser-cooled atoms in an optical cavity. The cavity both entangles the atoms and reads out their state, and a clock laser tracks the atoms as they oscillate at optical frequencies near 10 to the 15 per second. Vuletic described today's best atomic clocks as limited by atom quantum noise rather than technical noise, with fractional stability around 10 to the minus 18 rather than 10 to the minus 23; a clock at that level running since the Big Bang would be off by roughly 100 milliseconds.

At that stability, the clocks resolve gravitational redshift over a height difference of about 0.1 millimeter, a prediction that Einstein made and that satellite clocks confirmed decades earlier. Vuletic attributed the leading records in absolute spin squeezing to Mark Kasevich at Stanford and James Thompson in Boulder, with his own group's data running parallel to the Heisenberg limit about 10 dB below it.

Neutral-atom quantum computing and error correction

Neutral-atom quantum computing traps individual atoms in laser beams and uses their internal states as qubits, a platform Vuletic described as more recent than superconducting qubits and more directly scalable. Rather than fabricating a larger chip, the approach adds laser beams to trap, cool, and arrange atoms, which can be assembled into grids and other layouts.

Vuletic's group, working with Mikhail Lukin and Markus Greiner at Harvard, has demonstrated quantum error correction on this platform with results he described as similar in quality to those reported elsewhere. Error correction works by storing logical information non-locally in the entanglement between physical qubits, so a single stray photon does not destroy the encoded state. He said logical error rates of 10 to the minus 6 and beyond within the next couple of years look realistic, citing Hartmut Neven's presentation at the same event.

Black holes, error correction, and emergent spacetime

Quantum error correction, developed for quantum computing, gives physicists a language for black holes, and Daniel Harlow uses it to argue that spacetime itself is emergent rather than fundamental. In this picture, spacetime is the logical information encoded in a quantum error-correcting code, and the physical qubits live at its boundary.

Harlow framed the problem around three claims that cannot all be true. A black hole has a finite number of quantum states with an entropy of one quarter of its surface area in Planck units; its formation, evolution, and evaporation should be a unitary process; and an observer falling into a large black hole should see physics described by quantum field theory to all orders. Something has to give, and Harlow's work modifies the third: quantum field theory holds only for operations whose complexity is subexponential in the black hole's entropy.

What a 10 to the 5 logical qubit machine would run

A quantum simulator large enough to matter would test whether spacetime emerges from a proposed model of quantum gravity, according to Harlow. The workflow: someone proposes a theory, the simulator runs it, and the researcher checks whether spacetime appears instead of guessing. Harlow was explicit that this does not establish which theory describes our universe, only whether a given theory is internally consistent.

His proposed experiment involves reconstructing interior information from Hawking radiation at late times. The claim is that a sufficiently complex operation on the radiation can affect the black hole interior across a spacelike separation, an effect that classical spacetime structure forbids. Detecting it requires an operation of exponential complexity, tuned so the effect is large enough to observe and small enough to run. Harlow added a caveat the panel returned to: these results come from self-consistent models built to work this way, not from realistic theories.

Where quantum sensing and quantum computing meet

Quantum sensors already show a measurable advantage, and pairing them with small quantum computers is the most promising near-term combination the panel discussed. Vuletic called this a very promising avenue that nobody yet understands well, since a quantum computer is itself a sensor under the right conditions.

Mavalvala expects quantum algorithms to help gravitational-wave detection in a specific way. Gravitational waves themselves are classical, but the interferometer is a quantum instrument with over a thousand control loops, so the noise floor is quantum in origin. As technical noise falls, quantum algorithms become more useful for reaching lower-noise states, she said.

Harlow's answer to which scientific question comes next was blunt: quantum computers will transform science by simulating quantum systems we cannot solve, and he sees no other application that clearly beats that one. He also said no quantum error-correcting code yet does everything needed, noting that surface codes are paired with magic-state distillation and that newer low-density parity-check (LDPC) codes, which Jay Gambetta raised at the same event, took researchers by surprise. That surprise is evidence the field has not settled on its codes.

FAQ

• Is quantum advantage already real in sensing? Yes, in two documented cases. Squeezed light raised Advanced LIGO's sensitivity by 6 dB during the third observing run, and Vuletic's entangled atomic clocks report a 12 to 13 dB enhancement over unentangled measurements. Both are quantum advantages in metrology rather than in computing.

• What are quantum tools for scientific breakthroughs in practice? The working examples are squeezed states of light that lower interferometer noise, entangled atomic ensembles that beat the standard quantum limit in clocks, and neutral-atom platforms used to test quantum error correction. Quantum simulation of quantum gravity remains a goal rather than a delivered result.

• How many logical qubits would quantum gravity research need? Harlow asked for 10 to the 5 logical qubits before black hole simulations become informative, because realistic models have large numbers of degrees of freedom. Vuletic disputed that figure, citing known reductions in physical qubit counts for Shor's algorithm, and suggested it could fall further.

• Do quantum computers replace classical ones for this work? No. The panel described a hybrid architecture in which a quantum processor works alongside GPUs and CPUs, with computer science contributing to how the parts are interfaced. Classical control also stays essential for the optical and electronic systems that run the experiments.

• What limits quantum sensors today? Decoherence and optical loss, according to Mavalvala, plus technical and thermal noise that must be suppressed before quantum effects are visible at all. Vuletic noted that this is why superconducting qubits sit in dilution refrigerators and laser-cooled atoms run at microkelvin temperatures.

Source video