Latest · 1 min read
Say the quiet part to the model
A short companion to the longer piece on thinking out loud: the four things people leave out, every time.
Articles
Notes from the practice
Ideas, observations, and practical lessons about using AI to strengthen thinking, learning, and professional judgment. Explore how these ideas are being applied in classrooms, workplaces, and everyday decisions.
Archive
Most recent first
The homework problem is a design problem
Before you write a detection policy, check whether the assignment survives one honest question.
AI and assessmentA Tuesday afternoon audit
One page, eleven desks, four hours. The cheapest diagnostic I know for a team that has already bought the licences.
Capability at workThe Question Before the Prompt
Most people start typing before they know what they want. Here is a five-minute habit that makes every prompt afterward better — and it has almost nothing to do with AI.
Thinking partnershipThree questions I ask in every workshop
They take four minutes at the start of the day and they decide everything I do for the rest of it.
Thinking partnershipGrading in the Age of the Draft
If a student can generate a competent first draft in ten seconds, what exactly were we grading? A practical rewrite of three common assignments.
AI and assessmentPrompting Is Just Asking Well
Good prompts look a lot like good interview questions. Four moves borrowed from journalists, therapists, and eight-year-olds.
Prompting as inquiryCapability Beats Access
Every team I visit already has the tools. Almost none have the habits. What that gap looks like on an ordinary Tuesday afternoon.
Capability at workNotes from a room of skeptics
Forty-one teachers, one hostile opening question, and the three objections that turned out to be right.
Ethics and trustWhere I Draw the Line
Three tasks I hand over without hesitation, three I never will, and how I explain the difference to a room of skeptical teachers.
Ethics and trustThinking Out Loud, On Purpose
Talking to a model is closer to talking to a colleague at a whiteboard than to searching. Say the messy part first.
Thinking partnershipThe Rubric That Reads Itself
Building a rubric with students, then testing it against AI-written work. Step by step, including the part that failed.
AI and assessmentEnjoying these ideas?
Create a free account to access additional learning resources.
Create a free account →