ChatGPT
Turn a messy assignment or field note into a bounded brief, alternatives, and a verification list.
- FIRST RECEIPT
- A one-page brief with claims marked for checking
- WATCH
- Private, client, or regulated material
AI SERVICE DESK / FIELD KIT 003 / STUDENTS + RECENT GRADUATES
Pick something another person can inspect. Then choose the smallest set of services that can help you research, make, connect, ship, and explain it.
THE TOOL IS RENTED. THE JUDGMENT SHOULD BECOME YOURS.
01 / PICK A RECEIPT
Choose a mission. The desk will assemble four official doors, a 90-minute first move, and the proof to keep. Nothing here is uploaded or stored.
MISSION / PUBLISH
Turn one observed friction into a one-screen tool and put it on a public URL.
No account is connected. No action happens until you open an official door.
02 / OFFICIAL SERVICE DIRECTORY
These are starting points, not rankings. Pricing, access, regional availability, student offers, and model names change. Confirm the current terms on the linked service before committing money or sensitive material.
36 official doors
Turn a messy assignment or field note into a bounded brief, alternatives, and a verification list.
Read a long packet, challenge a plan, and make the assumptions and tradeoffs easier to inspect.
Ask questions of a source set you chose. Useful for lectures, papers, reports, interviews, and study packets.
Scout a topic on the open web, then open the citations and replace weak sources with primary ones.
Keep the source cabinet beside the AI. Save papers, metadata, notes, and citations you can inspect later.
Explore model cards, datasets, licenses, and Spaces before treating “open” as one simple category.
Give a coding agent the repository, constraints, acceptance criteria, and permission boundaries—then inspect the diff.
Use a second coding agent on the same bounded brief. Compare its first move, risk calls, tests, and finish.
Work inside a familiar code editor with repository context, suggested edits, and agent-assisted changes.
Use inline and agent-assisted coding inside the GitHub and editor workflow where your diffs already live.
Turn a compact app brief into a hosted prototype when local setup is the main thing stopping you.
Move from interface idea to a working visual prototype, then ask someone else to try the path without coaching.
Explore multi-reference, audio-video generation as a directed shot—not as a finished film.
Plan, create, and refine clips inside a filmmaking workspace built around Google’s generative models.
Generate and transform moving image with a toolset built for iterative creative production.
Generate and edit image, video, audio, and design assets inside Adobe’s creative workflow.
Build a visual language through controlled variation, reference images, and explicit keep/reject judgment.
Prototype voice, narration, dubbing, and sound only with voices and material you have permission to use.
Edit recorded audio and video through a transcript, then clean the story, captions, and timing before export.
Turn generated and captured material into an authored sequence through editing, sound, color, and export.
Make one server-side request, inspect the exact response, and build around the contract—not model memory.
Build a second model-backed path and compare output, latency, failure behavior, and cost on the same task.
Prototype a Gemini prompt and API behavior before moving the smallest useful contract into code.
Learn the common connection layer between an AI application and tools, data sources, or workflows.
Make the API call by hand before asking an agent to hide it inside an integration.
Connect an event to an action with visible steps, test events, logs, and code where the default action is not enough.
Prototype a no-code handoff across familiar apps; make the trigger, data path, and approval step explicit.
Add a real database, authentication, storage, or server function when the prototype needs durable state.
Make your process inspectable through repositories, issues, branches, commits, pull requests, and releases.
Check the current verified-student offers before paying for developer tools, domains, hosting, or learning products.
Put a static or full-stack project on a public URL from Git, direct upload, or the command line.
Deploy a web project with preview URLs and a Git-backed release loop.
Build a hosted checkout in test mode so payment state, success, cancellation, and fulfillment become observable.
Prototype SMS or voice with verified recipients, explicit consent, delivery status, and a clear stop path.
Send a transactional email from code while keeping templates, delivery events, and domain setup inspectable.
Instrument the one event that proves whether your public prototype is being used as intended.
No door matches that filter. Try the output you want—“video,” “deploy,” “citations,” “payment”—instead of a brand name.
03 / STUDENT WALLET RULES
A subscription is not progress. Buy the next constraint only when the artifact tells you what is missing.
Check the current GitHub Education pack and campus licenses while you still have verified eligibility.
Fake data, verified recipients, sandbox payments, and preview deployments are enough for the first proof.
Write the weekly credit or API budget before the first generation. Turn on usage alerts where the service offers them.
Your brief, source list, code, edit decisions, and receipt should survive if every paid service disappears tomorrow.
04 / THE CAMPUS ASSIGNMENT
Solo or in a group of three. Start with a person who does the work. End with a public-safe receipt.
Talk to one operator. Name one repeated friction and the evidence that would count as improvement.
SHIP / 1-PAGE BRIEFUse one coding or creative agent. Keep the interface narrow enough for another person to try today.
SHIP / WORKING DRAFTAdd one API, dataset, connector, payment sandbox, or messaging trial. Keep credentials and personal data out of the client.
SHIP / CONTRACT + TESTGive the artifact to one person without coaching. Record where they hesitate, what breaks, and what you will not fix yet.
SHIP / OBSERVER NOTEShow the live artifact, source or process, one failure, one decision, and the next honest step.
SHIP / URL + 60-SEC TOUR05 / HOW THIS DESK WORKS
Every outbound card points to the service’s own product page, documentation, or official education program. Industry Next does not receive affiliate compensation from these links.
“Start-free,” “student,” “open,” and “metered” are navigation lenses—not promises. Check current pricing, eligibility, data terms, regional access, and feature availability at the source.
SOURCE PASS / AUGUST 3, 2026 · REQUEST A CORRECTION: MH@POINTCAST.XYZ
Open the machine-readable directory