I started my day at the gym, checking my phone on short breaks and opening a LONG email about testing instructions. It had important directions. Before AI, I would not have read this until I missed important steps or deadlines! I copied and pasted the text into a private AI (Duck Duck Go), removed identifiable information and prompted, “Tell me what action items I need to know:”
- Dates to be aware of
- Steps to communicate with staff and
- Responsibilities to set up before testing.
When I arrived at my office, I noticed an email to approve a grant request requiring a three-sentence rationale expressing why I support the grant. I knew about the grant and the teacher’s explanation. I asked the AI chatbot to give me a three-sentence endorsement of the grant. The output produced deliberate information endorsing the grant. The concise response was nearly perfect (I removed the word “fully” from “fully endorse” as this sounded redundant, a common hindrance of AI outputs.
What else can you do with AI?
Next, I saw an amazing image on Google (OK, this wasn’t work, but I had to share). I took the image, dumped it into Gemini and asked what it was. Gemini identified it as the northernmost post for US space exploration in Greenland. Wow, that’s cool!
A scheduled observation conference with a teacher on my calendar reminded me to write the evidence for this (strong) lesson. When I am in a teacher’s classroom, I write anecdotal notes. Next, I ask the teacher to input their evidence into the online portal. I consolidate evidence and ask an AI to write it from the principal observing their lesson’s perspective.
I remove identifying information (student names) and set delimiters such as “rewrite this teacher’s reflection of their lesson in three sentences.” I proofread the rewritten text to ensure that I agree with it and edit the parts I don’t like since AI likes to use certain words that sound awkward.
I added anecdotal notes to the evaluation evidence — this is more analog-style — I’m not using AI; it’s my authentic words. However, the final stage involves me taking their evidence and my anecdotal notes, combining them into an AI and prompting: Based on the Danielson Model for Teacher Evaluation, take this evidence for the lesson and give me a five-sentence summary.
I re-read to confirm agreement with the overview. Next, I prompted the AI, which tracked all previous inputs, to provide suggestions for growth. This is probably my favorite part. Think about it: AI is pulling all my evidence and offering methods to further develop the lesson and feedback is relevant. I typically agree with most of this, edit a bit to add my literary voice and move on. Remember, this is deliberate evidence taken from what I gave the AI, so its trust level is high, and I have an expanded repertoire of suggestions to consider.
AI for administrative tasks frees up time
Wow, all that before lunch. What next? Here is how the real magic happens. I just spent the morning offloading many administrative tasks. That means I can be in classrooms, meeting teachers and engaging students and staff in the cafeteria, where important interactions happen, and I feel liberated from mundane work. My grant approval isn’t lingering because I had to pour over the detailed linguistics. My observation report is well-written and ready, and I even learned about that cool space agency in Greenland.
Then comes the afternoon email from a parent that is far too long and has what I need to know embedded in the seventh paragraph. You know the tone — highly emotional, too many CAPS and questionable accuracy. How do I dissect this? I remove all identifying information, plug it into a chatbot, and prompt it to prompt what I need to know in neat bullet points. It produces a beautiful summary, and I can respond accordingly.
The bullet points were spot on. This matters because I want to make sure I know the facts of the parent’s concerns, not every dramatic, nuanced, redundant detail. Remember, I can trust the AI because I am feeding it deliberate content (minus confidential content) rather than asking generic questions (how do you respond to a parent’s email complaining about a student’s issues?)
AI also helps solve problems fast
While digesting lunch, I receive a message from facilities, with a notice to inform the school community of a standard pesticide treatment. I find the scientifically written version unnecessarily confusing and loaded with jargon. I prompt with ChatGPT: reword this for my parent community.
This solves two problems fast:
- Eliminates the untrusting perception that we are using big words to confuse families about the treatment.
- Ensures I didn’t jumble my own words during an analog editing process. AI provides a grammatically correct response. Before AI, I was consumed with rewriting to make sense and meet linguistic conventions. Big time saver!
The day is almost over, and I received a request from a parent to send a text blast reminding families of a Saturday community park clean-up. Her message didn’t provide easy copy-and-paste text, so I had to limit it to 140 characters. I pasted the text and prompted: Rewrite this as a text in 140 characters. A cohesive message framed within the limits was produced. Imagine trying to rewrite it to make sense and fit 140 characters. Let AI do the work!
All of this took about an hour to complete. In my analog days, this would have taken hours or days. I keep thinking about the additional presence I give to students, parents and staff as a school leader. If AI can help me stay on top of the administrative minutiae and be more present, kids win. That’s all that matters.
Opinions expressed by SmartBrief contributors are their own.
Subscribe to SmartBrief’s FREE email newsletter to see the latest hot topics on edtech. It’s among SmartBrief’s more than 250 industry-focused newsletters.