Quantum simulation is a key scientific challenge and accurate large-scale simulation of quantum systems will be invaluable for understanding and engineering nature at the atomic scale. Quantum computers promise to revolutionize simulation by natively emulating quantum many-body systems. However, achieving practical computational advantages is challenging, since quantum processors are intrinsically noisy, and existing algorithms and prohibitively expensive. In this talk, I will introduce neutral atom arrays, and discuss recent theoretical and experimental progress towards performing useful quantum simulations. By co-designing Hamiltonian simulation algorithms and hardware-level control, we show how to programmably engineer a variety of model Hamiltonians. Then, we discuss how these techniques can be applied, in the near-term, to problems arising from materials science, chemistry, and high-energy physics. Finally, we present recent implementations of quantum error correction on this platform, and discuss a path towards achieving large-scale and accurate quantum simulations.