AI in 5: Group Projects with a Droid: AI as a Thought Partner in High School PBL (May 4, 2026)

What if the most productive member of every student group project... was an AI? In this episode of AI in 5, your AI Learning Guide JR unpacks how high school teachers can use AI as a structured thought partner in project-based learning — without turning it into a cheating shortcut.

You'll hear how a College Board study found 84% of high school students are already using generative AI for schoolwork, and why that's a signal to act strategically, not panic. JR walks through three practical classroom moves: AI-assisted idea generation with constraints, AI-powered project planning, and AI as a first-round feedback reviewer. Plus three student guardrails — Visibility, Transformation, and Attribution — that keep authentic learning at the center.

Backed by a 2025 Education Sciences study showing AI-enhanced PBL produces a large effect size (Cohen's d = 1.30) over traditional PBL, this episode gives you a concrete, one-step challenge to try before your next major project deadline.

Guest voices: Sal Khan, founder of Khan Academy · Dr. Jessica Howell, VP of Research, College Board

📚 APA CITATIONS (Show Notes)

College Board. (2025, October 6). New research: Majority of high school students use generative AI for schoolwork. College Board Newsroom. https://newsroom.collegeboard.org/new-research-majority-high-school-students-use-generative-ai-schoolwork

Khan, S. (2023, May). AI in the classroom: The world's best tutor [TED Talk]. TED Conferences. https://www.ted.com/talks/sal_khan_ai_in_the_classroom_the_world_s_best_tutor

Tariq, R., & colleagues. (2025). The role of artificial intelligence in project-based learning: Teacher perceptions and pedagogical implications. Education Sciences, 15(2), 150. https://doi.org/10.3390/educsci15020150

Transcript

SPEAKER_00 0:13

Picture this. You walk into your classroom during project work time. One group is sketching on a poster board. Another is arguing about who's actually going to present. And a third group is quietly huddled around a laptop, talking to an AI chatpot like it's their very own Portical Droid. So here's the question. What if the most reliable member of every group project isn't a student? It's an AI sidekick. Today on AI in 5 on the special May the 4th, we are going to explore one idea at the intersection of AI and education in under 5 minutes. I'm JR, your AI learning tour guide, and today we're looking at how AI can act as a thought partner in high school project-based learning without turning to the dark side of cheating. Segment number one. Oh great, the AI can do the whole mission for us. That's not what we want though. Instead of thinking as AI is a droid on a team, more R2D2 than Sith Lord, it's great at scanning possibilities, running diagnostics, and flagging problems. But it doesn't get to be the hero of the story. The students do. So the frame becomes AI is your co-pilot, not your pilot. It can help you brainstorm plan and stress test ideas, but the thinking and the final decisions must stay with the humans in the room. One line you can use with the students: AI can help you use the force of your own brain more effectively. It doesn't get to do your learning, though, for you. Let's make this more concrete. Imagine your students are working on an environmental science or civics project, design a plan to make our school more sustainable. Without AI, the usual pattern is they jump to hyperspace using Google, pull together some random ideas, solar panels, recycling, maybe rainwater collection, and try to bolt them into a slideshow at the last minute. With AI as your thought partner, you might have groups start with a prompt like, Act as a sustainable might work on this tiny campus. What did the AI miss because it doesn't know your local context? What assumptions is it making that it don't fit your reality? AI charts the star maps, but students still have to choose the course and fly the ship. First practical move. Idea generation with constraints. Build this into your project launch. Require every group to show you two things before they lock into a project. An AI generate a list of possible approaches. Their reasonings for choosing or rejecting each one. Student prompt might sound like we have three weeks, a class of 28 students, and almost no budget. Give us five project ideas to raise awareness about team mental health at our school. Now the learning happens and how they respond. They cross out ideas that are unrealistic, adapt promising ones, and at least one idea that came from them, not from the bot. You can walk around and ask very Jedi master-like questions like, why did you reject this one? What did you change here? Where do you think the AI was off target? AI is a holochron of possibilities. Students are the padawans who have to interpret and apply what they see. Second move, project planning and rule clarity. Group projects don't fall apart because a content is from a galaxy far, far away. They fall apart because no one knows who is doing what or when. Here, AI can act like a protocol droid with a clipboard. Have groups ask. Create a simple three-week project plan for four students. Include tasks, suggested deadlines, and who might be responsible for each. AI will spit out a basic timeline, more like a generic mission briefing. Then students personalize it, swap in real names for student A, tweak deadlines to match your calendar, add or remove tasks to fit the actual project. You can quickly skim these plans and see where crews are trying to blow up the Death Star with a water pistol, and what does have a realistic flight path. Third move, feedback and iteration. Instead of you being the one giving the feedback, AI can act as a first-round test audience. Students paste in a draft survey, script, or intro paragraph and ask, what might be confusing about this? What assumptions are we making? What question should we be ready for? Then require a short change log. What feedback did the AI give? What did we change because of it? What did we disagree with and why? Now AI is like the honest droid in the corner saying, the odds of this presentation landing with your human audience are very low. And students get a chance to improve before they face an actual human audience. You can still deliver the final verdict, but AI helps them rehearse. To keep AI firmly on the light side, you can set up three simple class rules. Rule one, visibility. If AI helped, it has to show up in your process. Ask groups to include one or two screenshots or copied snippets of AI conversations into their slides, notebook, or digital log. Rule two, transformation. Nothing goes into the final project exactly as the AI wrote it. Students have the paraphrase, combine, critique, or extend AI suggestions. No copy pasting from the droids. Rule 3, attribution. If I ask who came up with this idea, it's okay to say the AI suggested it and we decided to keep it because dot dot dot. In this classroom, hiding AI use is the dark side. Honest attribution is the way of the Jedi. So here's your one-step challenge. In your next major product, add just one explicit AI step, either brainstorming, planning, or first-round feedback. Then, after it's over, ask yourself, did this make my students more thoughtful and organized, or more passive? If it's the former, you've just recruited a new droid to your classroom crew, one that never skips class, never falls asleep in group work, and never gets the final say. And this has been your AIN5. This may the fourth be with you. And may your next group project be just a little bit more like a well-run rebel mission, and a little less like a council meeting in the Galactic Senate. If you've enjoyed this episode, pass it on to a friend, a colleague, parent, even a student to let them know what's out there. And of course, like us on all our social media on to AI innovations unleashed on pretty much everything. So until next week, be good, my Petavons. Thanks for listening. See you next time.