Quantum computing uses qubits, which require extensive preparation and measurement. Researchers at MIT's Engineering Quantum Systems Group (EQu S) explored using AI to streamline this process. Beatriz Yankelevich used GPT-5.6 Sol, integrated with Codex, to automate routine tasks in superconducting qubit experiments.

The AI successfully ran measurements on a six-qubit chip, saving significant time. However, it struggled with weak or noisy signals, needing human guidance. This shows AI can handle defined workflows but still faces challenges in ambiguous situations.