A language model is a machine-learning model trained on extensive text data to process and generate human language. Here, it becomes a practical source of realistic test data when production data does not yet exist. section 9.1.2
Shape generated test data so prototypes resemble the eventual product:
- Specify realistic columns, including dates, times, names, or addresses. section 9.1.2
- Describe realistic values and constraints, such as geography, future dates, or valid rating ranges. section 9.1.2
- Request realistic distributions, such as subscription spending or house prices, and refine the prompt when needed. section 9.1.2
Survey prototype data
ChatGPT generated two respondents answering three survey questions, returning respondent identifiers, question text, and numerical responses in a comma-separated structure. section 9.1.2


































































