While writing An Inbox Between Us, I ran experiments to see whether AI could recognize the patterns organizations leave behind in ordinary work: emails, meetings, handoffs, and recaps.
Since then, the models have changed noticeably. They respond more carefully, and anything approaching behavioral interpretation gets far more guardrails than it once did.
The patterns underneath did not change at all.
This week’s edition of The AI Mirror is about that gap. I rebuilt one of those early experiments inside today’s much narrower boundaries, and the constraints made the exercise more disciplined, not less useful.
The full prompt is included, along with a caution about what it is for and what it is not.
Note: The image is fake, however I do have a real one. Due to the nature of the information, I am not including my own personal version of this, as it is really for you and your own reflection of observable work dynamics.