Insider Brief Google researchers introduced AlphaQubit, an AI-powered decoder that improves quantum error correction, reducing errors by 6% compared to tensor networks and 30% compared to correlated matching. AlphaQubit’s two-stage training — pretraining on synthetic data and finetuning with experimental data — enables it to adapt to complex real-world noise, including cross-talk and leakage, showcasing machine learning’s potential in quantum computing. While AlphaQubit excels in accuracy, chall...