Bob and co-host Dr. Jessica White pick up the AI literacy conversation they started last episode, this time going further into how it changes what students learn and how teachers teach. They're joined by Rosie Giannetti, Assistant Director of Professional Learning and School Improvement at ACES, and returning guest Stacey Simpson, a professional learning specialist with a background in high school reading.

If you missed the previous episode, start there. It lays the groundwork: AI literacy is less about chasing the best tool and more about building thinking skills that hold up as the technology keeps changing.

Vulnerable learning vs durable learning Rosie draws the line between performance and understanding. A student can memorize that 6 times 7 is 42 and pass the quiz. Whether they can explain why multiplication works and apply it to something new is a different matter. Durable learning travels with you. Vulnerable learning falls apart the moment the context changes.

4 questions to test a task These came out of Jess's work on AI literacy and help teachers check whether a task asks for real thinking:

Does it require students to think, not only complete steps?

Does it require transfer to a new situation?

Could a student finish it successfully without understanding the content?

Does the evidence produced match the level of thinking you claim to assess?

Articulation is not thinking A line from Rosie's chapter: AI accelerated articulation, not thinking. AI can organize and express an idea cleanly. It can't supply reasoning that was never there. Bob sits with this one, noting we may be living through the first time a person can produce language without being the source of it.

Why AI can't replace the teacher Learning is relational. AI can deliver information, but it can't notice when a student is frustrated, build trust, create belonging, or convince a child they're capable of more than they believe. The group keeps returning to the calculator debate as the closest historical parallel.

Is this the end of writing? Maybe the opposite. The panel reframes the worry. The skill moves from the product to the prompt, and clear constraints and examples might make students stronger writers rather than weaker ones.

Stacey's nephew Max A pre-med freshman who uses AI as a thinking partner, not a shortcut. He's motivated to learn the content because he'll need it, and he leans on AI to get a second explanation when a lecture doesn't land. His story points to the value of making the why behind learning explicit, and to the trouble with policies that assume every student is trying to cheat.

What teachers are actually doing with it Jess describes the arc she watched over the year. Teachers moved from drafting parent emails toward designing better tasks, building custom gems, and using NotebookLM inside their PLCs. Rosie shares an aha moment with science teachers who redesigned a weak performance task into an authentic one. Stacey talks about matching struggling secondary readers with texts at their level that don't look elementary, so students stay engaged instead of shutting down.

AI as a mirror A thread running through the episode: technology reflects what we value and where our systems create barriers, including barriers to equity. The work is adjusting what the mirror shows, not getting mad at the mirror.

The time paradox Does AI give time back, or does the saved time just fill up with more work? Jess gives an honest answer. She isn't banking hours so much as going deeper, producing more specific and individualized work than she could before.

A first look at the ACES Curriculum Creator Jess previews the tool ACES built through vibe coding. It supports curriculum writing without doing it for the team, drawing on UBD, UDL, and Connecticut's design principles. Writing teams build courses, units, and lessons, and the platform generates lesson plans with facilitation notes, differentiation, student worksheets, and editable slide decks, plus admin tools for auditing vertical alignment.

Resources

• ACES Curriculum Creator: https://www.acespdsi.org/curriculum-creator

Bob and Dr. Jessica White welcome three ACES professional learning specialists to the show: Nicole Beauchamp, Melissa Rosenthal, and Stacey Simpson. Together they pull back the curtain on what's actually happening with AI in Connecticut classrooms, and the answers might surprise you.

Stacey shares findings from interviews with three students at different stages, from middle school to college. All three reported the same thing: their schools were banning AI outright, threatening zeros for anyone caught using it. Meanwhile, students like Stacey's daughter, who has dyslexia, are quietly teaching themselves to use AI as a learning partner. She breaks down word problems, parses confusing language, and accesses grade-level math without the shame that used to come with the struggle.

The conversation digs into the hard questions. Is AI helping students reach deeper understanding, or just helping them survive an inflexible system? Where should educators protect productive struggle, and where should they remove barriers? Nicole offers a memorable comparison to the shift from horse-drawn carriages to cars. The infrastructure takes time to catch up, and right now education sits squarely in the messy middle.

In this episode:

Why students across grade levels keep hearing "don't use AI" and what that messaging costs them

The shift in educator thinking from "how do we catch cheaters" to "how do we redesign assignments"

Universal scaffolds: how AI supports both literacy and math learners, including multilingual students and kids with executive function challenges

The CRAFT prompting method and why phased AI interactions beat one-shot answers

A look inside the ACES K-12 AI literacy curriculum, including second graders learning prompting concepts without ever touching a device

Middle school skills: verification, bias recognition, and understanding algorithms

High school as the driver's seat: student agency, systems design, and vibe coding in the new AI Foundations course

Practical first steps for teachers starting mid-stream, beginning with a simple class survey

About our guests:

Nicole Beauchamp is a professional learning specialist at ACES specializing in math, curriculum development, and building thinking classrooms. She has been in Education for 14 years- 12 as a high school math teacher at East Hartford High School in CT and 2 as an Instructional Coach in Bristol School District in CT. This is her first year as a Math Learning Specialist with ACES.

Melissa Rosenthal is a professional learning specialist working closely with the ACES Center for AI. A former reading interventionist and coach, she has led AI workshops for educators and administrators across Connecticut, including a statewide monthly alliance for districts building AI policy and implementation teams.

Stacey Simpson is a professional learning specialist with nearly 20 years of experience as a high school English teacher, reading interventionist, and instructional coach. She specializes in dyslexia identification and intervention.

Connect with ACES and the Center for AI to learn more about AI literacy workshops, the K-12 curriculum, and professional development for your district. https://www.ACESpdsi.org

Megan DaCosta teaches fourth grade math and science in Windsor, Connecticut. This is her first year bringing AI into the classroom in a real way, and she joined Bob and Jess to talk about what that looks like with 9 and 10-year-olds.

Her district runs on paper and pencil at the intermediate level, on purpose. So Megan's challenge isn't more screen time. It's making the technology she does use count. She works with Magic School AI, building chatbot rooms and station activities her students move through while she stays close.

The best moment of her year came from a mistake. A student practicing multi-digit multiplication got the wrong answer from a chatbot, and it argued with her. They solved the problem four different ways. When the student explained their reasoning, the chatbot backed down and admitted the student was right. The student still brings it up. They beat the robot, and the sudent knows it.

That's the AI literacy lesson hiding inside a math problem. Don't accept the first answer. Question it. Prove it.

In this episode:

• Using AI as a thought partner instead of a replacement for teaching
• Reaching 40-plus students a day without losing your read on each one
• Designing lessons backward so every skill level gets what it needs
• Teaching healthy skepticism without breeding cynicism
• Why she wants students to keep their creativity, not flatten it