The Morning Strategy Field Guide to AI

The Experiment

This morning, I came across a reference to three interesting systems thinking books on Reddit. I took a screenshot, uploaded it to ChatGPT, and asked for summaries.

‍

‍

The results were solid, concise, insightful, useful.

But then I had a different kind of thought:
What would these authors say about my work and things I'm questioning right now?‍

And a follow-up:
I don’t have those questions written down. So...?

‍

And I asked:
“Based on our chats, think of the kinds of questions I’m currently grappling with. For each question, suggest how these authors would answer.”

That's it - this simple prompt opened a door.

---

‍
ChatGPT scraped our previous conversations and surfaced four questions that were structured articulations of things I’ve been ruminating on, but never formally captured. It was like reading a distilled version of my own thought process, my active problem spaces. Only clearer.

Then, it responded to each question, responding as each book or author might. And I found myself at a table (desk) with three intellectual mentors, each offering a different lens on the same question.

‍

‍

The Reflection

This small experiment opened something big. It showed me how AI can:

• Surface the questions I didn’t know I was asking
  Most tools wait for inputs. This one mined the shape of my thinking and offered back some core unnamed tensions.
• Turn reading/research into active dialogue
  Rather than “What's this book about?”, I asked, “What would this book say to me?” That brought the content into my own context, brought it to life.
• Curate a perspectival, deliberative council
  I didn’t want a single answer. I wanted friction, contrast, convergence. The AI played out a deliberation andoffered a synthesis. A tiny rehearsal of systems-level decision-making.

‍

‍

This feels like a new form of sensemaking: something between a book club, a journaling practice, and a consulting session. It gives body to abstract concepts, then puts those embodied concepts in conversation with my actual work.

I’m imagining a custom GPT stocked with my favorite thinkers, frameworks, and questions, all ready to respond not with one answer, but three different ones, from three very different angles, converging into a possible path forward.

‍

But even this first, lightweight test was profound. Not because it gave me new ideas, but because it helped me recognize and name the ones already forming.