AI streamlines quantum computing experiments with GPT-5.6 sol

Researchers at MIT’s Engineering Quantum Systems Group (EQuS) are using GPT-5.6 Sol to automate routine measurements on quantum chips, allowing scientists to focus on more complex tasks. Graduate student Beatriz Yankelevich tested the AI’s ability to run and refine measurements on a six-qubit chip, which is commonly used for benchmarking fabrication processes. GPT-5.6 Sol autonomously selected measurement parameters, operated the hardware, and analyzed results, significantly reducing the time required for calibration. Yankelevich noted that while the AI excels at well-defined workflows, it struggles with ambiguous or noisy signals, often requiring guidance from experienced researchers. She emphasized that the AI’s ability to run measurements overnight or while she works in the cleanroom has freed up her time for higher-level tasks like interpreting results and planning experiments. "I’ve built infrastructure to guide agents through several parts of my work—measurement, theory, and chip design—and now it’s really starting to pay off," she said.