Ask
Start with a problem, curiosity, or outcome—not perfect technical language.
OpenAI Student Collective University of MiamiBUILD / EXPERIMENT / SHARE
Explore project prompts, see what is waiting to be built, and turn curiosity into something other students can try.
03 / THE ENERGY
Our best meetings end with twenty browser tabs open, a messy whiteboard, and something new existing that didn’t an hour ago.
01# Tonight at the Collective
02question: “Can an AI be creative?”
03experiment: build a tiny creative agent
04constraint: 45 minutes, teams of three
05result: show us what surprised you
06
BUILD / THE COLLECTIVE LOOP
If you can describe it, you can start building it. The first version only needs to be useful enough to test.
Start with a problem, curiosity, or outcome—not perfect technical language.
Add context, constraints, examples, and a clear definition of success.
Turn the idea into a first useful version you can actually interact with.
Give it to another student, observe what breaks, and improve the experience.
Demo the result and the lessons—not just the polished final screen.
BUILD / CHOOSE YOUR TOOL
You do not need to master everything. Choose the tool that matches the next step in your idea.
Explore ideas, understand material, write, research, plan, and turn scattered thoughts into a clear direction.
Try it for a study plan, research brief, or presentation outline.Bring files, project context, and repeatable workflows together to move a larger task from start to finish.
Try it for an event plan, application tracker, or team project.Describe what you want to make, then work with a coding agent to create, test, debug, and improve real software.
Try it for a website, campus tool, data project, or prototype.05 / IDEA GENERATOR
Pick a direction and get a starter prompt for the next collective experiment. These are sparks—not assignments.
What are you curious about?
BUILD / STARTER PROMPTS
Each prompt is a launchpad. Replace the brackets, add your perspective, and keep asking better questions as you build.
A useful student-facing guide or planning tool.
Help me prototype a simple tool for University of Miami students who need to [specific campus need]. First ask me five questions about the student, the moment they would use it, and what a successful result looks like. Then propose the smallest useful version we could build in one session.
A personalized, repeatable learning workflow.
Help me design a study system for [course]. I will provide the syllabus, learning goals, and my current notes. Organize the material into concepts, practice questions, and a weekly review rhythm. Clearly separate facts from anything that needs verification.
A testable one-page concept and interview plan.
Act as a rigorous product partner. Help me turn this startup idea—[idea]—into a one-page concept. Identify the user, painful moment, current alternative, riskiest assumption, and smallest experiment. Do not validate the idea automatically; challenge weak assumptions and end with five interview questions.
An interactive creative prototype with a point of view.
Help me create an interactive storytelling experience inspired by Miami. Before building, ask me to choose a medium, audience, emotional tone, and one constraint that makes the idea distinctive. Then produce a concise build plan and a first working version.
A sourced interface that helps others understand a topic.
Help me create an explorable research brief about [topic]. Define the key questions, propose trustworthy primary sources, and design a clear interface for claims, evidence, uncertainty, and open questions. Never invent citations, and flag any claim that still needs verification.
A working prototype that matches students to opportunities.
Help me design a student opportunity finder for [audience]. Begin by defining the minimum information needed from a student and from an opportunity. Then propose the matching logic, privacy considerations, empty states, and the smallest prototype we can test with five students.
06 / PROJECT QUEUE
These are starting points for our first cohort. Once members begin shipping, this space becomes a living showcase of real student work.
A practical, student-written guide to using AI thoughtfully for learning, making, and research.
OPEN FOR COLLABORATORS →One evening, small teams, and working prototypes centered on a problem students actually experience.
OPEN FOR COLLABORATORS →A growing collection of useful prompts, experiments, and lessons contributed by members across majors.
OPEN FOR COLLABORATORS →