Which AI tool, workflow or prompt is genuinely helping you make better decisions?

16 replies

That second question stopped me in my tracks. It is a brilliant signal in the noise.But it made me want to push the prompt one step further:What would change if the person most affected by this answer could actually see it?I ask this because I have lived the alternative. I know what it feels like to be the ghost in the room — the variable everyone talks around but nobody considers. It is an...

I use a version of your second question. I ask for any bias and the assumptions behind it. Keeps it simple for me and lets me see, very often, where I have made assumptions of it my own and had my own bias blinkers on. If I am feeling really nit picky, I am ask what decisions it has made for me. That one can be an eye openerIf I am making up resources I often so a "Is there anything you can...

The prompt that changed how I work:What is this response assuming about the person who will read it?It catches blind spots faster than any other question I know.

I create an anonymized catalogue of learning profiles and challenges for the students in each of my classes, then follow up the creation of a lesson or assignment with, "Apply the catalogue for Class X to this assignment. What challenges are likely to arise? For whom?" Helps me do the best I can in terms of UDL for all, plus offer bespoke scaffolding as necessary.

Using AI as a second lens on evidence, not as the decision-maker. For example, in my teaching role at Lytton High School, I find it useful to gather evidence first. Evidence from student work, assessment results, reflection notes, classroom observations, or feedback and then use an AI tool like Gemini to help surface patterns I may have missed. The prompt is usually something like:“Analyse...

One prompt I find surprisingly useful is: "What assumptions are we making here, and which of them might be wrong?"As an educator and school owner, I often use AI to help me think through decisions, design learning experiences, or explore new ideas. This question helps me move beyond the obvious answer and spot blind spots I may have missed. Sometimes the most valuable thing AI does isn't give...

I love this approach. It shifts the focus from "Is this a good lesson?" to "Is this a good lesson for these learners?" The anonymized catalogue is also a great reminder that personalization doesn't always require more work. Sometimes it just requires asking better questions about who might struggle, who might thrive, and who might otherwise go unnoticed.

Exactly, human agency must be the driver.

That distinction — talking about versus talking with — is one I carry into everything I build. The learner's voice is not just underrepresented. It is the signal we keep mistaking for noise.

Thank you Mr. Fitzpatrick — your second question is the one that refuses to leave the room.