The current capabilities of ChatGPT allow experienced practitioners to learn new techniques and to better organize and document their work, but raise issues of fair use, false attribution, and academic misconduct by some users and students. This paper analyzes the responses to three questions and its use to recognize fraudulent social media posts. ChatGPT gave an insightful and useful answer to an interpersonal relationship question. Its response to a technical question suggests that a skilled nonexpert can, with effort, produce a publication quality paper with little or no domain knowledge, and it would be difficult for a domain expert to detect the fraud. The response to the factual question illustrates an error also observed in some human work.