A priest wanted to test the speech-writing abilities of ChatGPT. He tested it in his own way, and received a text stating in general terms that it was Sunday and that’s why we are in the church. Everyone found this to be very weak, and it spread in his circle of acquaintances that ChatGPT cannot write a homily.
Then a believer came and brought him a sermon from ChatGPT, which the priest thought was very good. He even said that there was a Shakespeare quote in the text. Impressive, but how is it possible?
Well, this is prompt engineering knowledge.
Because anyone who can write such good text with a “machine” if he/she knows how to give the task. But only those who have an idea of what the text should contain can do it well. In our case, the following was needed for a good sermon:
- The text is written by a Catholic theology-trained priest.
- The topic is the Ten Commandments.
- There should be a relevant quote from Shakespeare.
- There should be a New Testament reference.
- The whole text should be metaphorical.
- The lesson should be that it is difficult to keep the commandments, but they give a healthy life.
What is the general lesson here?
Then it is not theft from ChatGPT if you know what parts should be included. If you know the purpose of the text. If you know how the individual parts should fit together. And it is the user’s text if he or she edits the whole thing himself/ herself. Not as a post-editor, let alone as a proofreader, but as the creator of the text.
This requires task breakdown, which computer programmers already know well.
Congratulations if you made it this far in the reading and thank you for your attention.
Leave a Reply