This paper presents a unified mathematical framework for continuous-time machine learning, which models temporal dynamics as a continuous process. It compares different branches of CTML, highlighting trade-offs and design choices, and identifies open challenges in implementation and research.