Output — What do you want to see from the AI when you're done?
"Give me a summary of key stakeholders, their positions, and 3 policy precedents from other cities"

Prompt — What instructions does the AI need to accomplish the task?
Be specific: tell the AI your role, your angle, and how to respond

Input — What context or documentation can you share with the AI?
Load the Bee U data pack: Andre's slides, impact report, survey data

Test — How will you know it's working?
Did it cite real sources? Does it match what Andre told you? What's missing?

Framework established in partnership with Decoded Futures

Bee University NYC Context Pack

Download these documents to use in your AI research today.

Open Context Pack

SPRINT 1

Set Up Your Starting Context

Save Your Context Pack

Create a folder on your computer called bee-university-research. Download all files from the context pack and save them there.

Define Your Research Focus

Open a new Google Doc or Notes file. Answer each prompt in 1-2 short bullet points. Save it as research-focus.

My Research Focus

- Track: (Policy / Environmental Impact / Career Pathways)
- The problem I want to explore is _____.
- What I don't know yet and want to find out is ______.
- My target audience for this research is _____
  (e.g. Bee U leadership, city council, funders, community members)
- "By Week 5, our team's work would be useful if it could help Bee U _____"

SPRINT 2

NotebookLM Deep Dive

Upload Docs to NotebookLM

Open NotebookLM with your CCNY Google account. Create a new notebook for your track. Upload the documents from the Bee U context pack.

Ask NotebookLM to Summarize Your Context

Start with this prompt, then refine based on what you get back:

I'm a college student researching urban beekeeping in New York City. Bee University NYC uses beekeeping as a gateway to engage young people in environmental issues and green jobs. There's an informal ban on beekeeping in NYC public parks that limits their reach. What can you tell me about this?

Go Deeper as a Team

Use AI as a brainstorm buddy. Try follow-up questions like:

"What evidence supports this?"
"What other resources might we consider?"
"What's missing from this data?"
"How does this compare to other cities?"

Try the Audio Studio

NotebookLM can generate audio overviews of your uploaded documents. Try it as a different way to absorb the information.

The Prompt Loop: Ask, Read, Refine

Your first prompt is never your best prompt. Here's how to iterate:

Prompt 1 — Start broad

I'm a college student researching urban beekeeping in New York City. Bee University NYC uses beekeeping as a gateway to engage young people in environmental issues and green jobs. There's an informal ban on beekeeping in NYC public parks that limits their reach. What can you tell me about this?

Prompt 2 — Add specificity

That's helpful but too broad. I'm specifically on the Policy track. Can you tell me: Who are the current NYC Parks borough commissioners? What is their known stance on allowing beekeeping in public parks? And are there examples of other major US cities that allow beekeeping in public parks that we could use as precedents?

Prompt 3 — Make it verifiable

For each city precedent, give me the specific policy name or regulation, when it was enacted, and a source where I can verify it. Flag anything you're not confident about.

AI Safety & Responsible Use

Three Rules for This Studio

Transparency — Document what AI did vs. what you did.
Verification — AI outputs are drafts, not facts. Always check citations and policy claims.
Proportionality — Use AI where it adds value. Your strategic judgment is the deliverable.

Your Team Task Before Next Friday

Do 30 minutes of AI-assisted research on your track as a team
Each team member tries NotebookLM individually with the context pack
Document 3 things you learned and 3 questions you still have
Come back next week ready to share your biggest finding

Resources