Google’s AI breakthrough brings quantum computing closer to real-world applications

Google researchers have discovered a new technique that could finally make quantum computing practical in real life, using artificial intelligence to solve one of science’s most enduring challenges: more stable states.

In a research paper published in NatureGoogle Deepmind scientists say their new AI system, AlphaQubit, has proven remarkably effective at correcting persistent errors that have long plagued quantum computers.

“Quantum computers have the potential to revolutionize drug discovery, materials design, and fundamental physics, if we can make them work reliably,” Google explains. announcement bed. But nothing is perfect: quantum systems are extraordinarily fragile. Even the slightest environmental interference – from heat, vibrations, electromagnetic fields or even cosmic rays – can disrupt their delicate quantum states, leading to errors that make calculations unreliable.

A walk research paper highlights the challenge: quantum computers only need an error rate of one in a trillion (10^-12) operations for practical use. However, current hardware has error rates between 10^-3 and 10^-2 per operation, making error correction crucial.

“Some problems that would take a conventional computer billions of years to solve would take a quantum computer only a few hours,” Google says. “However, these new processors are more prone to noise than conventional processors.”

“If we want to make quantum computers more reliable, especially at large scales, we need to accurately identify and correct these errors. »

Google’s new AI system, AlphaQubit, wants to solve this problem. The AI ​​system uses a sophisticated neural network architecture that has demonstrated unprecedented accuracy in identifying and correcting quantum errors, displaying 6% fewer errors than previous best methods in large-scale experiments and 30% fewer errors than traditional techniques.

It also maintained high precision on quantum systems ranging from 17 qubits to 241 qubits, suggesting that the approach could scale to the larger systems needed for practical quantum computing.

Under the hood

AlphaQubit uses a two-step approach to achieve its high precision.

The system first trains on simulated quantum noise data, learning general quantum error patterns, and then adapts to real quantum hardware using a limited amount of experimental data.

This approach allows AlphaQubit to handle the complex effects of real-world quantum noise, including interference between qubits, leaks (when qubits leave their computational state), and subtle correlations between different types of errors.

But don’t get too excited; you will soon no longer have a quantum computer in your garage.

Despite its accuracy, AlphaQubit still faces significant obstacles before its practical implementation. “Each coherence check in a fast superconducting quantum processor is measured a million times per second,” the researchers note. “Although AlphaQubit is excellent at accurately identifying errors, it is still too slow to correct errors in a superconducting processor in real time.”

“Training at larger code distances is more difficult because the examples are more complex and sample efficiency appears lower at larger distances,” a Deepmind spokesperson said. Decrypt“This is important because the error rate scales exponentially with code distance. So we expect to need to resolve larger distances to achieve the ultra-low error rates needed for fault-tolerant computing on large, deep quantum circuits.

Researchers are focusing on speed optimization, scalability, and integration as critical areas for future development.

AI and quantum computing form a synergistic relationship, enhancing each other’s potential. “We hope that AI/ML and quantum computing will remain complementary approaches to computing. AI can be applied in other areas to support the development of fault-tolerant quantum computers, such as calibration and compilation or design of algorithms,” the spokesperson said. Decrypt“At the same time, people are looking at applications of quantum ML for quantum data and, more speculatively, quantum ML algorithms on classical data.

This convergence could represent a crucial turning point in computer science. As quantum computers become more reliable through AI-assisted error correction, they could, in turn, contribute to the development of more sophisticated AI systems, creating a powerful feedback loop of technological progress .

The age of practical quantum computing, long promised but never realized, may finally be closer – but not close enough to start worrying about a cyborg apocalypse.

Edited by Sébastien Sinclair