If you have the API, you can track if the results change by setting temperature to 0, over time the difference to the same questions should not change drastical

If you have the API, you can track if the results change by setting temperature to 0, over time the difference to the same questions should not change drastically. I think this is the gold test, and not from using ChatGPT by feeling it out. The best place to track this would be to use GitHub