In this paper, the development of weather forecast technology is reviewed. Owing to people’s creativity and wisdom, the atmospheric evolution processes can be solved by modern mathematical and physical methods of model equations instead of merely judgement based on experience in the past. However, the complicated weather and climate systems still pose challenges to nonlinear uncertainty problems. The artificial intelligence (AI) technologies may contribute to address the challenges. This paper introduces some initial assumptions and development of AI in observation and identification, data processing, shortterm and nowcast forecasting, model parameterization and integrated analysis of forecast products and so on, indicating the way forward of further research and development and their contribution.