The Podcast
Why you should watch it ☝️
Derek Tranchina has spent 18 years in New Jersey public education. He taught middle school math for a decade, five years in an urban district and five in a suburban one, and then moved into tech coaching and professional development. Today he's Director of Special Projects at the Monmouth Ocean Educational Services Commission, a public nonprofit that provides services to school districts. His work there includes overseeing the New Jersey Virtual School and a GED program at a local jail. On the side, he consults with districts, and for the last three years almost all of that work has been about AI.
The idea I keep coming back to is Derek's answer to AI cheating. He compares generative AI to the calculator. Math teachers fought calculators too, and schools eventually stopped grading rote arithmetic and started grading number sense and reasoning. Schools had years to make that shift with the calculator, and ChatGPT gave them no time at all. Derek's conclusion is one a lot of teachers don't want to hear: if AI can complete an assignment quickly and easily, the assignment probably isn't a good one. He applies the same logic to the bar exam. If a model can pass the test but couldn't do the job, the test is measuring the wrong thing.
Derek is just as clear about what worries him. He points to Jonathan Haidt's The Anxious Generation and the rise in anxiety and depression among kids that started around 2010, the year Derek identifies as the arrival of the Chromebook, the front-facing phone camera, and Instagram. His answer is to use technology where it transforms learning and to make the rest of teaching social, hands-on, and human again.
Then there's Traffic Lite, the tool Derek built himself with AI. Many schools have adopted a traffic light policy for AI, where red means no AI on an assignment, yellow means limited use, and green means students can use it freely. Traffic Lite is a Chrome extension, free for teachers, that puts that policy inside Google Classroom. A teacher clicks a color, and the policy language goes into the assignment instructions where students and parents can read it. Administrators get a view of how teachers are applying the policy across a school. The product started as a student check-in tool until a friend saw Derek's post, assumed it was an AI policy tool, and asked for one. Derek built the product his first customer thought he already had.
Takeaways
If AI can complete an assignment quickly and easily, rewrite the assignment. Banning the tool won't hold, and the calculator already showed what does.
When a model can pass a test but couldn't do the job, the test is measuring the wrong thing.
A policy only works where people see it. Traffic Lite puts the AI rule in the assignment itself, in front of students and parents.
Students are more honest than we assume. Most of them want to know whether AI is allowed, and clear expectations do more than detection.
Pay attention when a customer misunderstands your product. Derek's best idea came from a friend who thought Traffic Lite already did something it didn't.
Change inside a rigid system depends on leadership. Derek credits a culture that answers "not yet" instead of "no."
Here’s a teaser…
Where to find Derek Tranchina 👇
On Linkedin.
What You Missed on Sunday
Here’s what we covered in Sunday’s newsletter edition…
AI Saved You an Hour. It Cost Me Two.
In April 2025, Shopify CEO Tobi Lütke told every employee that using AI was "a baseline expectation," and that anyone asking for more headcount or resources had to justify first why AI couldn't do the job. This month, the same CEO lamented that his employees were tossing AI "slop grenades" at each other.
Lütke hasn't changed his mind about AI. He's describing a cost his own mandate never planned for: work that took seconds to generate and hours for someone else to read.
This week: why AI made writing nearly free and reading more expensive than it's ever been, who ends up paying for it, and why company-wide AI mandates keep producing more of it.







