Meteorologists have used computer forecast models for decades to predict future weather conditions. These models use current weather observations and equations to generate forecasts. However, artificial intelligence (AI) operates differently. It relies on vast datasets of past weather events to teach itself what is likely to happen in the future. This difference gives computer scientists hope that AI can improve the accuracy and detail of weather forecasts, with longer lead times. AI helps extract more information from observations and formats it for input into weather forecast models. While research is exploring the possibility of replacing traditional models entirely with AI, it is currently unlikely.

That's because AI cannot predict weather events it hasn't seen before, such as record temperatures or strong storms. Additionally, AI produces forecasts pixel by pixel, whereas meteorologists are interested in weather systems and temperature patterns. However, there is one promising application: AI could greatly assist forecasters and fire crews in battling aggressive wildfires, providing high-resolution forecasts that account for fire spread. Nonetheless, it will take time and further evaluation before AI can be fully trusted in such high-risk situations. Collaborations between AI computer scientists, meteorologists, and social scientists are crucial for developing effective AI forecasts. Ultimately, the future of weather forecasts will likely involve a combination of traditional models and AI.