The Madden–Julian Oscillation (MJO) is a large-scale weather pattern in the tropics that moves eastward and can influence storms, rainfall, and temperature patterns around the world weeks in advance. Because of this, it plays an important role in improving weather and climate forecasts on subseasonal timescales (weeks to a month). However, scientists still don’t fully understand how far in advance the MJO can be predicted or what limits our ability to forecast it. This talk will summarize recent progress in understanding what makes the MJO predictable and where the main challenges remain. Key factors that help with prediction include knowing the right starting conditions and understanding how the ocean, land, and atmosphere interact. New methods using artificial intelligence and machine learning (AI/ML) show promise for improving forecasts, but there are still difficulties in making sure these tools are accurate, interpretable, and physically meaningful. Looking ahead, combining traditional weather models, theory, observational data, and AI/ML approaches may lead to better predictions—not only of the MJO itself, but also of related events like El Niño, tropical cyclones, and heatwaves. The talk will also highlight key challenges and suggest directions for future research.