Edge computing, as one of the key technologies in 6G networks, establishes a distributed computing environment by deploying computation and storage resources in proximity to end users. However, the dense deployment of base stations, cellular dead zones, and high dynamics of mobile devices may cause serious interference issues and weak signal propagation, which will severely affect the transmission efficiency of edge computing and cannot support low-latency applications and services. Reconfigurable intelligent surface (RIS) is a new technology that can enhance the spectral efficiency and suppress interference of wireless communication by adaptively configuring massive low-cost passive reflecting elements. In this article, we introduce RIS into edge computing to support low-latency applications, where edge computing can alleviate the heavy computation pressure of mobile devices with ubiquitously distributed computing resources, and RIS can enhance the quality of the wireless communication link by intelligently altering the radio propagation environment. To elaborate the effectiveness of RIS for edge computing, we then propose a deep-reinforcement-learning-based computation offloading scheme to minimize the total offloading latency of mobile devices. Numerical results indicate that the RIS-aided scheme can improve wireless communication data rate and reduce task execution latency.