News and Updates
Improved Chatbots, Research, Paul on Decoder Podcast
HUGE Updates to the big models
In early December, OpenAI (ChatGPT), Google (Gemini), and Anthropic (Claude) all released major improvements and new capabilities to their chatbots. We’ll dig into the practical uses of these new features next month, but here are the highlights:
ChatGPT 5.2 offers a ton of new connections, including FREE access to Adobe products (like Photoshop) inside of the ChatGPT apps
Gemini 3 brings Nano Banana Pro to all users with amazing photo editing capabilities. Like having a robot that can do Photoshop edits for you.
Group chats in ChatGPT allow multiple users to interact simultaneously with the AI
ChatGPT Plus offered free to educators
New Study looks at Student AI Use
A professor at UMass-Amherst let half his economics students use AI tools like ChatGPT during the semester. Students with AI access felt more confident and engaged but did not score higher on exams. The study shows AI can help students learn more efficiently without improving test results.
Paul Shared Perspective on Decoder Podcast
In its recent episode about how educators view AI tools, the Decoder podcast (hosted by The Verge’s Editor-in-Chief, Nilay Patel) featured teacher voices including our very own Paul Cancellieri. While Paul’s take was more optimistic than that of the other educators that they spoke with, the entire episode is a great listen.
AI-Supported Lesson of the Month
Celebrating Students More Easily With AI
Any classroom teacher who has struggled with classroom management has probably learned that the trick is often to spend at least 5x more time on positive feedback than on correcting negative behaviors. Investing in showcasing the good will usually pay dividends in minimizing the less desirable habits.
What’s the Goal? Celebrate students with a “Student of the Week” program driven by student feedback
How do I get started?
I designed a simple Google Form that asked students to nominate their classmates and provide justification for their choices. I intentionally made the form anonymous to minimize the scrubbing of Personally Identifying Information before handing off the task to the AI tool.
After replacing student names with numbers and removing any mentions of other students, I created a Notebook in Google’s NotebookLM for the project. From there it was simple to ask the model to use the spreadsheet data to help me choose a Student of the Week (SOTW) and write a short blurb about them. I proofread the output, made some changes, and asked the tool to help me generate graphics for my classroom slides.
Going forward, each week I just need to update the spreadsheet and remind the model that I don’t want to repeat any students. The Notebook keeps track of previous winners and helps make this habit more manageable for me too. Students love being recognized and I enjoy their engagement on SOTW days.
Have ideas for how to tweak this system and make it more effective? Share your spin with us!
Upcoming Talks and Appearances
Where is Paul this month?
The holiday season brings less travel and more writing for Paul. The Learning Forge draft is coming along and this spring’s conference presentations should help hone the message that it will share. Paul has some availability in January to come to your learning space if your team is ready to get started, or needs help moving to the next phase of your AI development. Reach out today!
That’s it for this month.
Stay tuned for more examples of lessons from Paul, and from the amazing teachers that he gets to learn with around the USA. Paul will be presenting at the Science for All Summit in February and the NCTIES conference in March. If you’ll be at either of those events, please reach out!
Paul (and the Codium Educational Consulting team)
P.S.
Don’t forget to email Paul with examples of how you’ve used the tools and strategies that he shared with you.

