“The three-dimensional world of ordinary experience—the universe filled with galaxies, stars, planets, houses, boulders, and people—is a hologram, an image of reality coded on a distant two-dimensional surface.” Leonard Susskind’s words, uttered close to three decades ago, ring out still in the corridors of theoretical physics. Now, the quest for this vision of holography is no longer limited to blackboards and mathematics abstracted from the world. It’s playing out in the quantum computers’ circuitry and deep neural networks’ algorithms as physicists try to shed light on what is at the center of a black hole.
Central to that pursuit is the holographic principle, a revolutionary concept that reality as we know it gravity, space, and time and all could be a projection from an underlying lower-dimensional quantum system. This was a principle that arose to answer a paradox: the entropy of a black hole, an indicator of the disorder inside it, varies not with the volume but with the area of the event horizon. As Susskind and Gerard ‘t Hooft made precise, the actual degrees of freedom are inscribed on a surface, not in the volume itself. In 1997, Juan Maldacena provided this concept with mathematical form via the AdS/CFT correspondence, demonstrating that a gravity theory in a higher-dimensional “bulk” is dual to a quantum field theory without gravity on its boundary dimension.
This dualism is more than a curiosity of mathematics. As Enrico Rinaldi, a physicist at the University of Michigan and RIKEN, describes:“In Einstein’s General Relativity, space-time exists but there are no particles,” Rinaldi explains. “In the Standard Model, particles exist, but there’s no gravity.” The problem is how to reconcile these two landmarks of contemporary physics a quantum theory of gravity.
Rinaldi and his colleagues have looked to quantum computing and deep learning to investigate this bridge. Their target: quantum matrix models, mathematical representations that capture both the particle and gravitational properties of black holes. In string theory, these matrices are stacks of vibrating strings, and their ground state the lowest energy state could contain the blueprint for the architecture of space-time itself.
But to pull out the ground state from such models is a hard computational task. As Rinaldi describes, “It’s really important to understand what this ground state looks like, because then you can create things from it,” Rinaldi says. “So for a material, knowing the ground state is like knowing, for example, if it’s a conductor, or if it’s a super conductor, or if it’s really strong, or if it’s weak. But finding this ground state among all the possible states is quite a difficult task. That’s why we are using these numerical methods.” His analogy is suggestive: suppose the numbers in a matrix are grains of sand. If the sand is completely flat, you’ve achieved the ground state. But ripples high energy states need to be ironed out, and this becomes rapidly impractical for classical computers as the size of the system increases.
Step in the variational quantum eigensolver (VQE), a quantum-classical hybrid algorithm specifically designed for NISQ devices. VQE exploits quantum circuits strings of quantum gates applied to qubits to generate trial wavefunctions, whose parameters are progressively improved by a classical computer to achieve minimal energy for the system. This scenario is especially appealing for quantum simulations of many-body systems since it avoids the exponential growth condemning classical algorithms.
However, the road is not without hindrances. Existing quantum hardware is only capable of a few dozen qubits, and circuit depth is restricted by decoherence and noise. As the size of the matrix model increases, so does the number of qubits and the complexity of the quantum circuit to accurately model the entire ground state wavefunction. “Other methods people typically use can find the energy of the ground state but not the entire structure of the wave function,” Rinaldi said. “We have shown how to get the full information about the ground state using these new emerging technologies, quantum computers and deep learning.”
To counter these hardware limitations, scientists are crafting sophisticated variational methods. As an illustration, cluster-based VQE algorithms partition qubits into clusters according to their shared information, enabling larger systems to be simulated by dispersing the computation among smaller, less entangled circuits. Besides decreasing circuit depth and the number of qubits needed, this strategy also improves tolerance to noise a highly valued benefit for NISQ-era machines.
Complementing quantum circuits, deep learning methods notably neural network quantum states offer a powerful alternative for representing complex wavefunctions. By training neural networks to approximate the ground state, researchers can capture correlations that would otherwise require prohibitively large quantum circuits. In Rinaldi’s work, a neural network’s parameters are optimized to “level the grains of sand,” converging on the ground state configuration of the matrix model.
The implications are not limited to black holes. The holographic principle, if confirmed, would redefine the nature of the universe itself. Several theorists propose that space-time and gravity are emergent effects, consequences of more basic quantum rules inscribed upon a lower-dimensional surface. According to this understanding, the fabric of reality is woven from the entanglement of quantum bits qubits upon a cosmic screen.
More recent breakthroughs in quantum simulation for high-energy physics highlight the wider promise of these methods. Quantum algorithms are being designed to simulate lattice gauge theories, simulate Hawking radiation, and study the dynamics of black hole evaporation, usually in simplified toy systems to avoid the constraints of present hardware. The AdS/CFT correspondence, the heart of the holographic approach, gives a mathematical scheme for converting gravitational issues into soluble quantum systems.
Even so, the trip is not yet done. As Rinaldi and co-workers forge ahead toward more complicated matrix models and increasing qubit numbers, there remain the never-ending problems of noise, decoherence, and barren plateaus in optimization landscapes. Error mitigation techniques, hardware-efficient ansatzes, and hybrid quantum-classical paradigms are all research frontiers, each inching closer to a quantum simulation of gravity beyond the bounds of classical computation.
The potential of this work is deep. With the advent of quantum computing, deep learning, and the holographic principle, physicists are no longer peering beyond the event horizon they are mapping out the quantum architecture of space-time itself. As Susskind’s hologram becomes increasingly clear, the lines between information, geometry, and reality become ever more intertwined.
