Quantum technology promises to revolutionize how we learn about the physical world. An experiment that processes quantum data with a quantum computer could have substantial advantages over conventional experiments in which quantum states are measured and outcomes are processed with a classical computer. We proved that quantum machines could learn from exponentially fewer experiments than the number required by conventional experiments. This exponential advantage is shown for predicting properties of physical systems, performing quantum principal component analysis, and learning about physical dynamics. Furthermore, the quantum resources needed for achieving an exponential advantage are quite modest in some cases. Conducting experiments with 40 superconducting qubits and 1300 quantum gates, we demonstrated that a substantial quantum advantage is possible with today?s quantum processors. There is considerable interest in extending the recent success of quantum computers in outperforming their conventional classical counterparts (quantum advantage) from some model mathematical problems to more meaningful tasks. Huang et al. show how manipulating multiple quantum states can provide an exponential advantage over classical processing of measurements of single-quantum states for certain learning tasks. These include predicting properties of physical systems, performing quantum principal component analysis on noisy states, and learning approximate models of physical dynamics (see the Perspective by Dunjko). In their proof-of-principle experiments using up to 40 qubits on a Google Sycamore quantum processor, the authors achieved almost four orders of magnitude of reduction in the required number of experiments over the best-known classical lower bounds. ?YS Quantum-enhanced strategies can provide a dramatic performance boost in learning useful information from quantum experiments.