AI in the Classroom

In this episode we explore the idea of cognitive surrender in AI education: what happens when students, teachers, or professionals begin relying on AI not just for support, but for the thinking itself.

How do we use AI as a thinking partner without gradually losing our own voice, judgment, and generative ability?

Topics covered:

• What “cognitive surrender” means in the context of AI and education
• Why AI dependence is harder to detect than plagiarism
• The difference between recognizing your thinking and producing it independently
• How AI can subtly reshape voice, judgment, and creative work
• What teachers should consider before using AI for lesson plans, feedback, or writing support
• Why districts should pay attention to the defaults being set by AI tools
• How schools can encourage intentional AI use rather than reflexive dependence

Source:

https://www.oneusefulthing.org/p/choosing-to-stay-human

In this episode we explore a sharper way to think about AI in literacy instruction: not as a shortcut that makes texts easier for students, but as a teacher’s assistant that helps educators identify where complex texts may challenge readers.

Topics covered:

• Tim Shanahan’s argument for using AI as a teacher-facing tool
• How AI can identify complex sentences and vocabulary in classroom texts
• The difference between useful AI analysis and noisy or overly technical output
• Why teacher judgment still matters when interpreting AI suggestions
• What AI handles well in text analysis — and where it struggles
• Why text complexity is more than a single readability score
• What this means for teachers choosing and teaching complex texts
  

Sources:

https://www.shanahanonliteracy.com/blog/ai-and-the-teaching-of-reading-comprehension-part-2

https://app.newsela.com/view/cmlldktks00083b6qi19cz2vu/?levelId=cmlldktor0kt708ncdns10m6r&activityPanel=annotations

In this episode we explore what eighth-grade students in California are asking about AI. We reflect on a recent classroom visit and use students’ own questions to examine what young people already understand about AI, what they are unsure about, and why their concerns about fairness, cheating, fake work, and feedback matter for educators and district leaders.

Topics covered:

• How students are experiencing AI feedback in real classrooms
• The limits and risks of AI detection tools
• What student questions reveal about AI literacy
• Why transparency matters when AI gives feedback
• How AI use can affect students’ sense of fairness
• What teachers and districts should consider before adopting AI tools
• Why process, not just final writing products, matters in assessment

In this episode we explore the debate around Khanmigo, Khan Academy’s AI tutor, and what its struggles reveal about the difference between explaining content and actually supporting learning.

Topics covered:

• Why Khanmigo was described as “a non-event for most students”
• The difference between a great explainer and a true tutor
• How Sal Khan learns
• Why students often need more than prompts and explanations
• What AI tutoring tools may misunderstand about motivation and learning
• How teachers can evaluate whether an AI tool supports real student thinking
• What instructional coaches should help teachers notice about AI products
• What district leaders should ask vendors before buying AI tutoring tools

Sources:

https://punyamishra.com/2026/04/16/why-sal-khant-on-learning-by-making-but-teaching-by-telling/

https://substack.com/inbox/post/197857852

In this episode we examine a published analysis comparing civilian casualties at Gettysburg with civilian casualties in Gaza in 2024, and explain why the comparison is historically incoherent. 

Topics covered:

• Why AI-generated comparisons can sound plausible while being misleading
• The difference between factual accuracy and content-area coherence
• What the Gettysburg/Gaza analogy gets wrong
• Why teacher subject-matter knowledge is essential in an AI-rich classroom
• How AI-generated “slop” can pass through editorial and instructional review
• What classroom teachers should look for before using AI-generated materials
• Why instructional coaches need content-area reviewers
• What district leaders should ask vendors about educator review processes

Sources:

https://www.techlearning.com/technology/ai/in-an-ai-classroom-content-knowledge-matters-more-than-ever

https://www.edsurge.com/news/2026-01-12-ai-is-changing-classrooms-teacher-expertise-still-sets-the-direction

In this episode we explore why AI is increasingly showing up in education as a tool for coaching and feedback.

We look at Renaissance Philanthropy’s discussion of the Gates Foundation’s AI initiative, which is exploring whether AI can help support teacher coaching in primary education in sub-Saharan Africa.

Topics covered:

• Why teacher coaching and feedback are emerging as a major role for AI in education
• What the Gates Foundation initiative is trying to explore in sub-Saharan Africa
• Why “better than nothing” is a powerful but risky standard for AI tools
• What AI coaching could mean for teachers, instructional coaches, and district leaders

Sources:

https://www.renaissancephilanthropy.org/insights/embedding-ai-where-it-matters-teacher-support-and-instructional-practice-in-sub-saharan-africa

https://www.amazon.com/Innovators-Solution-Creating-Sustaining-Successful/dp/1664444459

In this episode we explore a federal lawsuit involving Palo Alto Unified School District, a student essay flagged by Turnitin’s AI detector, and the much larger question underneath the case: what should schools actually do when an AI-detection tool raises a red flag?

Topics covered:

• The Palo Alto lawsuit over a student essay flagged as AI-generated
• The workload problem for teachers managing repeated AI flags
• False positives and the limits of tools like Turnitin’s AI detector
• Why paper-only rewrites can create equity and accessibility concerns
• What instructional coaches can do to create shared schoolwide expectations
• How district leaders can stress-test their academic integrity process
• The difference between investigating authorship and grading student performance

Sources:

https://www.paloaltoonline.com/palo-alto-schools/2026/05/11/parent-sues-palo-alto-school-district-over-artificial-intelligence-procedures/

https://www.vanderbilt.edu/brightspace/2023/08/16/guidance-on-ai-detection-and-why-were-disabling-turnitins-ai-detector/

https://news.mit.edu/2026/study-ai-chatbots-provide-less-accurate-information-vulnerable-users-0219

In this episode we explore four practical questions educators and district leaders can ask before adopting AI tools in classrooms. Drawing on Daniel Willingham’s framework for evaluating educational technology, we look at how schools can move beyond “we can’t be left behind” thinking and make more deliberate decisions about when AI helps learning—and when it may quietly replace something important.

Topics covered:

• Why AI adoption is different from earlier classroom technologies
• Daniel Willingham’s four questions for evaluating new edtech
• The risks of adopting AI tools too quickly
• Why “narrow” AI tools may be easier to evaluate than broad chatbots
• What schools may sacrifice when AI replaces older classroom practices
• Why districts need clear evaluation plans for AI initiatives
• How teachers and leaders can think more honestly about AI trade-offs

Sources:

https://www.facebook.com/DTWillingham/posts/i-still-hear-this-justification-re-ai-in-classrooms-its-out-there-we-cant-be-lef/1245588610883142/

https://theeducationequation.org/episodes/daniel-willingham-uva/transcript

In this episode we explore what it might take to design writing assignments that students are less likely to outsource to AI.

We look at a college philosophy assignment called “My Apologies” and use it as a lens for thinking about AI-resistant assignment design—not because every teacher should copy it, but because it reveals some useful principles for K–12 classrooms.

Topics covered:

• Why many traditional writing assignments may not survive easy access to AI
• The difference between preventing AI use and designing work students want to own
• What the “My Apologies” assignment asks students to do
• How AI changes the cost of generic assignments
• Why strong assignment design can make grading more meaningful for teachers
• Practical questions teachers can ask when redesigning writing tasks
• Why AI-resistant work is often more humane pedagogy

Sources:

https://philife.nd.edu/assets/308458/fullsize/drexler_apology_assignment.pdf

https://www.amazon.com/Good-Life-Method-Reasoning-Questions/dp/1984880322/

In this episode we explore what happens when AI helps students produce “better” writing, but makes that writing feel less like their own.

Drawing on a recent essay by Nurul Hassan Mohammad about AI use in STEM classrooms, we look at a subtle but important risk: when students use AI to polish lab reports, methods sections, and research writing, the work may become clearer and more polished while also becoming more generic. Are students losing opportunities to develop a disciplinary voice, to sound, think, and write like emerging scientists?

Topics covered:

• Why STEM writing is more than just a final write-up
• How AI-polished writing can flatten student voice
• Why “better” writing is not always deeper learning
• What student discomfort with AI-generated prose can teach us
• How AI may quietly reshape what counts as credible academic writing
• What teachers can praise, question, and protect in student writing
  

Source:

https://theconversation.com/college-students-are-noticing-their-ai-smoothed-writing-sounds-strong-and-not-like-them-279436