Artificial data created by computer for test purposes and machine learning. The more data given to a neural network in the training phase, the more effective its results when used in the production phase. However, synthetic data is generally not as useful as real data in blog posts, articles and books written by humans. See
AI model collapse and
neural network.