video1.0<iframe src="https://www.loom.com/embed/28a1cf73f54b4a01b4f25cb3aea6a193" frameborder="0" width="1920" height="1440" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>14401920Loomhttps://www.loom.com14401920https://cdn.loom.com/sessions/thumbnails/28a1cf73f54b4a01b4f25cb3aea6a193-d946f9ed5b97dbc1.gif1082.833I walk through how I set up system instructions and preferences for ChatGPT, like being direct, challenging assumptions, citing sources, and avoiding hallucinations. Then I show how I use connectors to my database for analysis, like answering enrollment questions for a class and spotting that enrollment is slowly declining. Next I download a public data set from data.gov, and I recommend using CSV or XLS for exploratory data analysis. I also asked ChatGPT to act as an expert CFO to price a program for at least a 40 percent profit margin, and I suggested checking for missing costs. No other viewer action was explicitly requested.