Brain first — then the machine

A two-part music video that teaches six principles for using generative AI well at university. Watch it, then see how each line maps to policy.

Part 1 — read the brief, and think before the machine (about 2 minutes). Captions available.

Part 2 — own it, check it, every tool, and ethics (about 2 minutes). Watch the full version.

The six messages

  1. Read the brief.The assignment brief decides what AI use is allowed. Live assessments are always unassisted.
  2. Let it support your thinking, not replace it.Think first, then use AI for ideas, explanation and counterarguments.
  3. Understand and own what you submit.If you can't explain it, don't hand it in.
  4. Check it against real sources.AI sounds confident but can be wrong or invent references.
  5. The rules cover all computer help.Not just chatbots: spell check, stats software and websites too.
  6. Data needs ethical approval.No AI analysis of research data without it, and only as participants agreed.

Brain first, then the machine,
Read the brief and know what it means,
Own it, check it, keep it clean —
Brain first, then the machine.

Annotated lyrics

Left: the lyric. Right: what it means in practice. Based on a university School of Psychology generative AI policy (2025); always follow your own institution's policy and assignment brief.

Verse 1 — Read the brief

Open up the module page, Read the brief before you type
Check the assignment brief before using AI at all.
If it says no, then leave it there, If it's silent, coursework's fine.
The brief overrides the general policy. Where it says nothing, submitted coursework may use computer help — understood, checked and cited.
But in the exam, in the room, On the stage with eyes on you, There's no search bar and no prompt
Exams, presentations and assessed conversations are completed without assistance.

Chorus

Brain first, then the machine
Don't make AI the first step in your thinking.
Own it, check it, keep it clean
Own it: don't submit what you don't understand. Check it: errors, bias, gaps. Keep it clean: cite your AI use.

Verse 2 — Support your thinking

Don't hand it the question first, Give your own mind the first go
Don't paste the exact question you've been set straight into AI.
Then let it argue back at you, Ask it what you'd like to know.
Good uses: ideas, explanation, feedback, counterarguments, generating questions.
And when they tell you to reflect — That's a mirror, not a guide.
Reflection is yours. Getting AI to reflect for you bypasses the learning.

Verse 3 — Understand and own it

If you couldn't talk it through, Then it isn't yours to hand in — It's pretending to be you.
Submitting work you don't understand is a breach; you may be asked to talk it through.
Looking like you learned it all When you didn't crosses the line
Using AI to give the impression you've learned more than you have is academic misconduct.

Bridge — Check real sources

It talks like it's certain, it sounds like it knows, But references vanish wherever it goes
AI output can be incomplete, misleading or biased, and references can be invented.
So find the real paper, the page and the line
Check against independent sources — ideally the original journal article.

Verse 4 — Every tool, and ethics

It's not just the chatbot, it's every tool, The spell check, the stats pack — the same honest rule.
The policy covers all computer assistance, on any device. Declare what you used.
And data from people is trust that they've lent: Get ethics approval, use it how they consent.
No AI analysis of research data without ethical approval — and only in the way participants agreed.

How this was made: brain first, then the machine

“You’re lecturing us about AI… using AI?” Yes, deliberately. This video follows its own rules:

  1. Read the brief.It’s built on the School’s AI policy, which I wrote. I checked every lyric against it.
  2. Brain first.I decided what mattered: I ranked the six messages and settled where the School and University policies differ. AI helped me think, not decide.
  3. Own it.Claude drafted lyrics; I checked, edited and approved every line. I can explain all of it.
  4. Check it.I checked the lyrics against the policy, and the timings against the song by ear, line by line, until they were right.
  5. Every tool counts.Suno generated the music and Claude Code built the animation, both credited openly, as the policy asks.
  6. Ethics.No personal or participant data was used.

AI did a lot of the making. The thinking, judgement and responsibility stayed with me. That’s the point.

Credits and reuse

By
Piers Fleming and Claude
Music
Generated with Suno
Lyrics
Drafted with Claude (Anthropic), edited by Piers Fleming
Animation
Built with Claude Code and Remotion
Context
Made for psychology students and staff at the University of East Anglia, autumn 2026.
Reuse
Educators are welcome to show it with credit.