Researchers at MIT Lincoln Laboratory have released a new artificial intelligence dataset called TorNet, which contains radar returns from thousands of tornadoes that have occurred in the United States over the past 10 years. The dataset aims to improve tornado detection and forecasting by enabling breakthroughs in machine learning algorithms. Along with the dataset, the researchers have also released models that have been trained on it, showing promise in their ability to detect tornadoes. Improved tornado detection and forecasting could help issue more accurate warnings, potentially saving lives. The dataset provides a benchmark for researchers and makes meteorology more accessible to data scientists.