Key points:
Most project-based learning workshops are built around three domains: design, assessment, and implementation.
In a model we developed for the Buck Institute for Education (PBLWorks) almost 25 years ago, day 1 of the workshop focused on project design, day 2 on project assessment, and day 3 on project implementation.
One of the key features of project implementation is the suggestion to launch a project with a Need to Know activity. The goal of this activity is clear: Every student should leave with an understanding of what they need to know and what they need to do to successfully complete the inquiry and generate a meaningful solution.
In practice, no matter how skilled the teacher is or how smart and talented the students are, many learners simply cannot transition from the launch to a concrete list of actions that will allow them to complete their tasks.
I recently saw a post on a discussion board that offered a solution, though it was targeted at office workers. The author shared a prompt that he uses after uploading a project description and expected outcomes to a chatbot: “I’m about to start this project. Interview me until you have 95% about what I actually want, not what I think I should want.”
That got me thinking. I wonder if there was a series of activities that allow a student to access the ideation capabilities of AI to ensure that they understand the challenge before them and recognize effective ways to utilize their skills and interests to complete it?
Reimagining “Need to Know:” AI-powered launch strategies
While the Need to Know list is a classic for identifying knowledge gaps, AI can act as a Socratic mirror, reflecting a student’s latent interests back to them until they recognize a personal connection to the driving question.
Here are five activities that you can try with your students to ease the challenge of getting started with their inquiry. You will notice that the focus here is on individual student work–not the group work typically found in PBL classrooms.
While these protocols are designed for individual work, they can be adapted for collaborative tasks. Teams can input combined interests, draft ideas, or early questions, then use AI-generated prompts to structure discussion. The key shift is that students respond first as individuals then negotiate meaning as a group.
1. The adversarial interest interview
Students engage AI as a skeptical questioner who challenges why a topic should matter.
- Sample prompt: “I am starting a project on [TOPIC]. I want you to act as a skeptical journalist. Ask me one challenging question at a time about why this topic should matter to me or my community. Do not give suggestions or ideas. Only ask questions that push me to clarify what I genuinely care about. Continue until I arrive at a specific angle that feels meaningful.”
2. Interest mapping & pattern extraction
Students input past experiences, interests, and frustrations; AI identifies themes and follows up.
- Sample prompt: “Here is a list of my past experiences, interests, and frustrations: [LIST]. Analyze this list and identify 3–5 patterns or themes you notice. Then ask me 5 follow-up questions to help me clarify which of these I care about most. Do not suggest a project topic.”
3. Contradiction finder
Students surface competing interests or values; AI highlights tensions and prompts reconciliation.
- Sample prompt:“Here are some things I’m interested in or care about: [LIST]. Identify any tensions or contradictions between them. Then ask me questions to help me explore how these conflicting interests might connect in a meaningful way. Help me think through the tension but don’t resolve it for me.”
4. Cross-domain collision
Students connect a personal passion to the academic topic through AI-generated “what if” scenarios.
- Sample prompt: “My project topic is [ACADEMIC TOPIC], and one of my personal interests is [HOBBY/PASSION]. Generate 3 ‘what Ii’ scenarios that connect these in unexpected ways. For each scenario, briefly explain the connection. Then ask me which one I’m most curious about and why.”
5. Scenario stress test (Need to Know Generator)
AI places students in a high-stakes scenario tied to the project.
- Sample prompt: “Create a realistic scenario where I am [ROLE] dealing with [PROJECT-RELATED CHALLENGE]. Give me 2–3 difficult decisions to make. After I respond, tell me what information I was missing that would have helped me make a better decision. Help me turn those gaps into a ‘Need to Know’ list.”
Final thoughts
I began this blog with reference to a prompt that focused on a project launch. The exchange that resulted from the prompt determined the worker’s understanding of the task and helped identify the skills and interests she brought to the process.
I want to flip the use of this prompt and make it a closing activity.
Here is a template for a prompt that could generate a final reflection rich in metacognition:
“I just finished the presentation of learning for my project on [TOPIC]. I am uploading the project description and the work products I generated [VIDEO/LINKS/DOCS/URL/PHOTOS]. Interview me until you can identify 95% of what I learned during this project, including the skills I developed (critical thinking, creativity, collaboration, communication, etc.) I developed. I am interested in learning more about my areas of strength and my opportunities for growth.”
If the original Need to Know helped students answer, “What do I need to know and do to complete this project?”, these AI-supported protocols push toward a more essential question: “Why does this work matter to me?” The shift may be subtle, but it is consequential.
In an AI-rich classroom where ideas are abundant and answers are cheap, the scarce resource is not information. It is ownership. When students use AI to interrogate their interests, test their assumptions, and refine their questions, they are not outsourcing thinking. They are making their thinking visible. That, ultimately, is the goal of any strong project launch.
Discussion about this post