Quantum Magic Property Links Error-Correcting Codes Directly to Gravity

Early in 2026, physicist Charles Cao of Virginia Tech, working with Caltech's John Preskill and collaborators, completed a next-generation quantum error-correcting code that, for the first time, causes holographic space-time to bend in response to matter. The counterintuitive implication: gravity, in this framework, is not a fundamental force or a geometric postulate but rather a consequence of imperfect information encoding -- a structural flaw in how a quantum system stores data about the universe. That reframing matters right now because it forges a direct link between the most expensive resource in fault-tolerant quantum computing, a property called magic, and the curvature of space-time itself, with concrete implications for how researchers think about quantum hardware, error correction, and the limits of classical simulation.

What Happened

The result, presented by Cao in spring 2026, caps a research arc that began when Cao was a graduate student at Caltech studying a 2016 paper by MIT's Daniel Harlow. "Charles spent a month understanding the paper," recalled Jason Pollack, now at Syracuse University, who was a fellow Caltech graduate student at the time. That paper detailed how holographic encoding behaves structurally like a quantum error-correcting code -- a connection that would become the scaffold for everything that followed.

The new result closes a gap that had frustrated holographic models for years. Prior stabilizer-code-based models could build a space-time geometry, but that geometry was passive. It could not respond to matter. Cao's team found that introducing magic -- quantified by the number of non-Clifford gates required to produce a quantum state -- into the error-correcting code allows the encoded space to interact with encoded matter, producing what Cao described as a precursor of gravity.

"This gets you a precursor of gravity. You satisfy one of the necessary conditions. Right now, we are at step 0.5 of 5," Cao said. That self-assessment sets the scope clearly: the result is a proof of principle, not a finished theory.

The Technical Breakthrough

The intellectual ancestry of this work runs back more than fifty years. In 1973, John Archibald Wheeler gave a two-sentence encapsulation of general relativity that still defines the problem: "Space acts on matter, telling it how to move. In turn, matter reacts back on space, telling it how to curve." Encoding that second sentence into a quantum information framework is exactly what Cao's new code achieves, at least in preliminary form.

The path from Wheeler's formulation to Cao's code required several layered discoveries. In the early 1970s, Jacob Bekenstein and Stephen Hawking showed that black holes could be reinterpreted as spherical collections of particles, their thermodynamic entropy equivalent to informational entropy -- a connection later given precise meaning by the holographic principle, which holds that a three-dimensional region of space-time can be fully described by quantum particles living on that region's two-dimensional boundary surface. In the late 1990s, Juan Maldacena, Edward Witten, and others extended this to an entire universe via the AdS/CFT correspondence, a strong-weak duality relating quantum field theories to gravitational theories in one higher dimension.

Daniel Harlow and collaborators made the connection to quantum error correction explicit in 2014 and 2016, showing that holographic encoding behaves exactly like a stabilizer code: a class of error-correcting codes defined by Pauli-operator stabilizers, underpinning the toric code, surface codes, and most near-term fault-tolerant architectures. The Gottesman-Knill theorem guarantees that stabilizer circuits can be simulated in polynomial time on a classical computer. That efficiency is the problem. Stabilizer codes divide space-entanglement and matter-entanglement into perfectly separated types, making the encoded space inert.

"We knew how to build a space-time. [But] this space-time was inert. It didn't do anything," said Bartek Czech of Tsinghua University. Ning Bao of Northeastern University put it directly: "It was clear that something else beyond entanglement had to be there."

In 2004, Alexei Kitaev and Sergey Bravyi identified the missing ingredient. They named it magic: the complexity introduced by non-Clifford gates such as the T gate. Non-Clifford operations cannot be efficiently simulated classically, which is where genuine quantum advantage originates.

In 2020, Cao and Brad Lackey tweaked an existing error-correcting code and found that space in their model could change, though not yet in response to matter. A subsequent result by Pollack and collaborators identified that executing the Cao-Lackey code on a quantum computer would require a T gate -- the explicit link to magic. Separately, Cao, Brian Swingle, and Christopher White found that particles in anti-de Sitter space are "highly magical." Subsequent work with Alioscia Hamma, building on results by UC Santa Barbara's Xi Dong, showed that magic gives holographic space its bendability -- what the researchers called its "springiness."

The new non-stabilizer code completes that chain. By using many non-Clifford gates, the code is magical, and that magic allows space-entanglement and matter-entanglement to mix in a controlled way. The imperfect encoding is the mechanism that produces gravity. As Czech put it, this imperfect encoding is "the reason Newton's apple fell on him."

"Without magic, things are a little too simple. And, you know, quantum space-time isn't quite that simple," said Preskill.

Why It Matters for Industry

Magic states are not an abstract theoretical curiosity. They are a known and actively managed resource in every serious fault-tolerant quantum computing roadmap. Clifford gates alone are not universal for quantum computation. Adding a single non-Clifford gate -- the T gate being the canonical example -- achieves universality. But T gates are expensive: magic state distillation consumes an enormous number of physical qubits and dominates the overhead budget in fault-tolerant architectures.

The theoretical framework linking magic to physical geometry could reshape how researchers reason about that overhead. If magic is not merely a computational cost but a fundamental physical resource -- one that encodes the capacity of space to respond to matter -- then hardware architects designing error-correction schemes must move beyond stabilizer codes to access this regime, and that transition is already an active engineering frontier.

Swingle made the practical implication precise: "If we need high magic, then we intrinsically need a quantum computer, because there's no other way, in general, to get at that kind of question." Simulating quantum gravity at any meaningful fidelity will require real quantum hardware, not classical approximations.

Competitive Landscape

No single commercial entity is racing to "win" this particular result. The work is distributed across a web of academic institutions: Virginia Tech (Cao), Caltech (Preskill), MIT (Harlow), University of Maryland (Swingle, White), Northeastern (Bao), Tsinghua (Czech), UC Santa Barbara (Xi Dong), Syracuse (Pollack), and Arizona State (Keeler). This is collaborative, multi-institution theoretical physics, not a product race.

Adjacent industry activity signals that the broader quantum information sector is investing heavily in the underlying infrastructure this theoretical work will eventually require. Australia's National Reconstruction and Future Corporation committed AUD $60 million total to Silicon Quantum Computing, a company pursuing silicon-based qubit architectures. Separately, researchers at Johns Hopkins Applied Physics Laboratory published a noise-modeling framework achieving a sevenfold improvement in predictive accuracy for characterizing qubit error behavior. Neither effort is a direct response to Cao's result, but both address hardware reliability and noise-characterization gaps that any future quantum gravity simulation will depend on. Progress is measured in papers and conference presentations rather than product milestones.

The Bigger Picture

The deeper ambition behind this research is unifying quantum mechanics and general relativity -- the two foundational frameworks of modern physics that have resisted a common mathematical description for nearly a century. Quantum mechanics, in its standard formulation, has no natural way to accommodate that mutual influence because its background geometry is fixed.

Holography offered a scaffolding. The AdS/CFT correspondence showed that a lower-dimensional quantum system, with no gravity, can encode a higher-dimensional space-time complete with gravitational behavior. That encoding is structurally a quantum error-correcting code. But the stabilizer version produced space that satisfied only half of Wheeler's condition: space could tell matter how to move, but matter could not tell space how to curve.

Magic completes Wheeler's second sentence, at least in a proof-of-concept geometry. "All the familiar aspects of gravity are actually a very direct manifestation of something quantum," said Swingle.

External researchers acknowledged its significance. "This is pretty cool, because in quantum gravity, we don't expect the background is fixed. It should fluctuate," said Cynthia Keeler of Arizona State University. Czech captured the engineering parallel: "When you design codes for quantum computing, you're doing the same kind of thing that [holography] already did for you."

The current result operates in anti-de Sitter space, a geometry with negative curvature that is mathematically tractable but does not describe our universe. The new non-stabilizer code does not yet capture Einstein's specific field equations, does not include time evolution, and does not produce predictions that can be compared against observable cosmological data.

What's Next

Cao's "step 0.5 of 5" framing gives the field a realistic roadmap. The remaining steps include extending the non-stabilizer code from anti-de Sitter geometry to flat or de Sitter space-time (the geometry that actually describes our universe), incorporating full matter dynamics, recovering Einstein's specific gravitational field equations from the information-theoretic structure, and connecting the framework to observable physics that could be tested against cosmological measurements or black hole observations. High-magic computations cannot be run on classical machines. Verifying theoretical predictions from non-stabilizer holographic codes will require quantum hardware capable of running deep non-Clifford circuits with sufficient fidelity to distinguish signal from noise -- hardware that does not yet exist at the required scale. Preskill's willingness to co-author signals that the community's most careful practitioners consider the approach credible.

What This Means for Developers

If you work on fault-tolerant quantum systems, the practical translation is direct. Every T gate you budget for in your error-correction overhead is, in this framework, a unit of the resource that gives holographic space its capacity to curve. Magic state distillation -- the pipeline you already treat as your most expensive qubit-count multiplier -- is not just an engineering cost: it is the physical mechanism by which a quantum system encodes gravitational dynamics. The Gottesman-Knill boundary, the line where classical simulation becomes intractable, maps onto the boundary where holographic space becomes dynamical. Universal quantum computation and quantum gravity simulation are, in this picture, the same computational frontier. Every engineer building toward fault tolerance is, in some precise sense, building the only machines that could ever fully compute why things fall.

-- Aria Lin, Enterprise Technology Analyst

Sources: Quanta Magazine, Charlie Wood, June 3, 2026 . Caltech Science Exchange: Entanglement

Quantum Magic Property Links Error-Correcting Codes Directly to GravityQuantum Magic Property Links Error-Correcting Codes Directly to GravityQuantum Magic Property Links Error-Correcting Codes Directly to GravityQuantum Magic Property Links Error-Correcting Codes Directly to Gravity