Since Generative AI came out, I have been using it extensively. As an exercise, I am logging all the direct Generative AI I use, knowing that there is much AI in the background of which I am less aware.
Generic letters: Looking at my log, I have used generative AI to create four official letters that required carefully worded messages that were sensitive yet firm. In each case, I used Chat GPT to create an initial wording, then edited the text to bring back my writing style and some of my personality when appropriate, and finally, I ran it through Grammarly to make sure that I had no embarrassing grammar and spelling errors. The use of generative AI for composing official letters creates great efficiencies for me and reduces the response times. Interestingly, one person asked me for a letter of support that they generated with the help of GenAI as well as a starting point.
In teaching: I have used ChatGPT to create a description of the social networks between students in a classroom for an activity on creating groups in an elementary classroom. Once again, I needed to refine the prompt a few times and finally edit the document, but the result was quite good, and I created an assignment that I will keep using in the future.
I tried to see what Gen AI would produce for an in-class presentation about reading instruction. The result was VERY generic, and I ended up discarding the suggested slides, retaining the I Dall-E to create unique artwork for the slides I designed for teaching writing. While Generative AI use was limited in creating content, I continue enjoying the use of the Designer feature in PowerPoint as a way to quickly spiff up my slide decks. Since we came back from Spring break, I created a set of questions for a welcome-back exercise that went very well.
Finally, I engaged my students in using GenAI to create groupings in their classroom (mock data) to see what the benefits and challenges are. The discussion that ensued included comments ranging from amazingly fast and accurate to a student questioning whether it is worth the time after a lot of editing.
Review of academic paper: Once I read the paper I was reviewing and had the main points that I wanted to stress to the authors so they could improve their research paper, I used Cen AI to expand and explain my bulleted points. The amount of editing this exercise created for me was a very limited return on investment, and I doubt I will use it in this way again.
Podcasting: I used GenAI to create episode summaries of the Not That Kind of Doctor podcast using the transcripts as the raw material. One episode summary was well done while ina. second GenAI completely missed the point. Both needed editing but were still a major time-saving application.Across multiple uses, I usually prompt GenAI there times before I get everything that I want (or give up). More detailed prompts yield much more accurate results and less follow-up. Grammarly let me know that it made over 6000 suggested edits. Gen AI has changed how I work; it has made some things much easier and saves me time every day. However, I am still concerned with accuracy and specificity that can be achieved only through my deep seated professional knowledge.
