What should we do with near-term quantum computers?

Soon, quantum computing researchers can expect to have access to a quantum computer consisting of 50 to 100 qubits. While fault-tolerant quantum computer promises great speedups over a range of important problems, these devices will not be reliable enough to fulfill these promises. An outstanding challenge then is to come up with an application that can reliably carry out a nontrivial task of interest. Can we do anything that is practically useful? I will argue that an important near-term application of a quantum computer would be the prediction of materialistic properties of challenging lattice models. This must be possible because quantum computer can significantly speed up the existing variational calculations for such models, and because the outcome of the computation is resistant to noise. I will report on an ongoing progress, and discuss the future prospect of this program.