Welcome new subscribers! As a reminder, this email is always free and will arrive in your inbox in the middle of every month. I’ll never sell you email address to anyone else, or send you any additional messages. But reach out to me if you have questions or want to dive deeper into any of the topics I share in these monthly newsletters.

Yesterday, Claude’s parent company Anthropic made big news announcing a new free tier of access to their Pro plan for educators. Teachers need to verify their full-time teaching status through a nebulous process that includes uploading documents and waiting up to several days. But successful enrollment unlocks access Cowork and Code, two very powerful features that many teachers (and most other users) have never used before. This creates some hurdles for school districts that need to ensure student data privacy and monitor appropriate use of AI tools by employees. But it may create new demand for these frontier capabilities that most folks don’t fully appreciate because they haven’t used them before.

At the annual ISTE conference a couple weeks ago, Google announced the integration of a new Classroom app into Gemini that can securely access class context streamlining the process of teachers using Gemini to create teaching materials relevant to their unique needs. Teachers will be able to create Guided Learning experiences, adaptive notebooks, and assessments; all while tracking student progress in Google Classroom. It looks like the connection between Gemini and Classroom is getting closer and that’s a good thing for teachers that want to use Gemini to help them build learning experiences.

A new law from the Illinois state legislature this month prohibits evaluators from using AI to assign numerical scores or qualitative rankings for components of a teacher evaluation. By reinforcing professional judgment and human expertise, the law is trying to draw a clear line between AI-assisted paperwork and AI-made judgments. This is going to spread quickly to other states, and I’m confident we’ll see more of this type of legislation soon.

Continuing last month’s share (The CAMPE Checklist), I’m going to pull another idea from my upcoming book, “The Learning Forge”. It’s called Defend the Delta and it is a series of activities that share one goal: Require students to critically evaluate the suggestions made to them by AI and their subsequent choices.

In mathematics and science, the symbol Delta (𝚫) represents change. In the Learning Forge, the "Delta" is the exact difference between your original thought and the AI's suggestion.

The Core Rule: If you can't defend the change, you can't submit the work. Recent cognitive science research shows that when we use AI to help us write or solve problems, our brains easily slip into "metacognitive laziness". We stop thinking critically and start passively accepting whatever the AI tells us. To avoid this, you must act as the Artisan. The AI can suggest changes, but you must defend why those changes make your work better.

While any work to require students to reflect on the AI-created work and assess its quality is super important, there are a few specific activities that I use with students for this purpose.

At the simplest end of the scale, I conduct personal conversations with students while they are working on assignments that may involve AI support. I ask them about what they’ve like and disliked about the output of the models. Often, they are initially surprised because so many of them don’t actually read the AI text before pasting it into an assignment. Just encouraging them to read it over is a great first step.

When your students have become comfortable with the habit of reading over AI-generated text, it’s time to put them in discernment mode. I impress upon my students the value of discernment as a superpower for the AI Age. Those people who can successfully evaluate the quality of what their AI models are giving them, and then sift the good from the bad, will do well. To help them with this task, I provide opportunities for students to categorize the sentences and ideas in their AI responses by their accuracy and usefulness.

When I’m ready for someone more formal—beyond reflection questions tacked onto an assignment or verbal discussions during work time—I use this document as a template for students to document and explain their editorial decisions with AI content. I make it clear that this type of decision-making isn’t optional—it’s required for students to be allowed to use AI to help them think and learn. And then I grade their analysis using the rubric below:

It really depends on the personality of your class and what motivates them to learn. For some of my students, making the case that changing AI-generated text preserves their unique human voice is enough to get them to do it. They value what makes them different from others and they don’t want the “endlessly average machine” to reduce those qualities of themselves and their work.

For other students, though, the drive to complete assignments and earn scores/grades is so profound that I need to tap into a different motivation to help them. These are the students that care more about the completion of the work than the underlying learning. Of course, I strive to help them escape from this rat race mentality, but in the meantime I use some trickery in the form of Defend the Delta.

I tell them how important it is to modify and personalize AI output because that’s how you outsmart the AI detectors and the adults that use them. Studies have shown that with a minimum amount of change, AI-generated content slips right past even the best detectors. This probably says more about the uselessness of these tools than anything else.

Want to see more? Check out the Learning Forge One-Pager for more tips from the upcoming book.

Writing has been the primary focus of the first half of my summer, but I’ve also been lucky to bounce around North Carolina on day trips to present about the Learning Forge. I keynoted a session for the Digital Technology & Computer Science Pathway of the NC CTE Conference in Winston-Salem, and then worked with Beginning Teachers at NC State’s PACKed BT Summit in Raleigh.

Next up, I’ll be giving a keynote at the “Abacus to Algorithms” workshop sponsored by The Friday Institute and the National Endowment for the Humanities. In August, I’ll be at the Kenan Fellows Program’s first-ever conference entitled “Reigniting Teacher Passion”. If you’ll be at any of these, I’d love to say hi and find out how you’re using AI as an educator!

Look for August’s newsletter to include more useful resources from the Learning Forge!

Paul (and the Codium Educational Consulting team)

P.S.

Don’t forget to email me with examples of how you’ve used the tools and strategies that I’ve shared.