Modeling the interplay between epidemics and human behavior has long been a complex challenge in computational epidemiology. Traditional models, such as the well-known SIR model, struggle to capture the dynamic feedback loop between behavioral responses and disease spread. In this talk, I will explore how the COVID-19 pandemic has fundamentally transformed our approach to this problem, largely due to the unprecedented availability of data. I will present findings from various epidemic-behavior models across different predictive tasks, examining their performance and drawing insights for future research and preparedness.