Lately I have been checking my work by discussing it with ChatGPT. I suspect that the lessons I am drawing might not be typical LinkedIn lessons on how to write better prompts. They are a good grounding in how NLP can be used to scaffold critical thinking. It’s definitely useful to receive immediate feedback on writing and strategy. But whether you are receiving feedback from a person or an AI, it pays to find a balance between defending your position and digging in your heels.
- As a writer, I expect that editing will mean letting go of some words or approaches and refining others. When I feed ChatGPT a new piece of work, sometimes it is a part of an ongoing conversation and sometimes it is only partially framed. This is like working with other people. Sometimes they are clear on what you want to achieve, and sometimes not. This changes how effective both the analysis and the suggestions are. The lesson is not that I should have fed more into the prompt. The lesson is that I must be alert to the frames that I can observe the editor using and correct those frames. At least ChatGPT doesn’t mind being told that it has not understood my objective.
- ChatGPT can be right in identifying weak spots and still give bad advice about how to correct them. If I were to take the suggestions as offered, often it would be at the expense of the meaning I want to convey and the rhythm of my language.
- ChatGPT offers a great reminder of how easy it is to be sucked in by compliments and rapport. My goal in using ChatGPT is to see my work from another point of view. When it is closely mirroring me, it makes it harder to achieve that goal–and that feels good. It’s nice to feel that the AI gets me. It’s not useful. Useful is finding the places where I haven’t achieved my objective so that I can think about how to do better.
If you’re using ChatGPT, critical thinking is required. It is a machine that quickly adopts or shifts frameworks and points of reference. It’s important to recognize that the voice coming from the machine has been programmed to make you feel good (rapport) and to start from conventional thinking that might or might not be useful to you. As it attempts to correct your writing, you might notice that it can’t quite catch your voice. That’s a good thing. It’s as important to find the edge to rapport as it is to create rapport.
Unlike your human colleagues, the program is not offended if you challenge its position or its strategy. This is a also a good thing. If your objective is to produce better work, you will need to challenge.
