My final goodbye to AI ragebait
Tilly Norwood's AI-made music video, "Take The Lead," is the kind of glossy slop that gets creatives foaming at the mouth. Dead-eyed visuals. Lyrics that read like a brand deck. It's engineered to poke your nerves and farm attention.
I felt that punch of frustration, too. Then I realized I'd handed over exactly what it wanted-my time, my takes, my reach. That's the trap. Outrage is the product.
The sound of slop
The song leans on choir swells, safe chord progressions, and a neat little arc about AI being "the key." It gestures at "human spark" while the avatar's migrating freckles and glassy stare remind you there's nothing behind the eyes.
Worse, it tries to have it both ways: "still human" while pitching AI as a superior path to scale and success. That tension-claiming humanity while dodging the cost of it-is why this kind of work feels hollow.
Why this stuff spreads
It's disingenuous on purpose. Promise a future, deliver a stock template, and let the internet do the rest. Media piles on, people dunk, and the cycle boosts reach for free.
Ragebait wins because we boost it. Every quote-tweet is an ad buy you paid with your day.
Spot the slop (so you don't make it)
- Uncanny faces: floating freckles, lifeless eyes, perfect skin with zero micro-movements.
- Emotion by shortcut: choirs, crescendos, and clichés in place of story and lived detail.
- Scarcity pitch in the lyrics: "don't be left out," "don't fall behind," "we can scale, we can grow."
- "I'm still human," but no credits, no data sources, no accountable process.
A better playbook for creatives
- Set your ethics in writing: consent, credit, compensation. Add content provenance via the C2PA standard.
- Publish your process: what was AI-generated, what was shot or performed, and who got paid.
- Use AI for what it's good at: research, ideation, variants, rough passes. Do the human bits that add texture: voice, reference gathering, location audio, real footage, edits with taste.
- Test small. Share WIPs with a trusted circle before you post it to feed the machine.
- Measure craft, not reach. Virality isn't proof of quality-nor is outrage.
How to respond to AI slop without feeding it
- Don't repost the video. Screenshots or timestamps if you must critique.
- Keep it useful: call out exact frames or lines and explain why they fail.
- Ship a better alternative. Recut a 10-15 second segment with your hybrid process and share the breakdown.
A practical hybrid workflow for music/visual projects
- Brief: define the one-line premise, emotion, and audience. Build a tight reference board.
- Guardrails: confirm rights for datasets, models, and references. No likeness theft, no shady scraping.
- Pre-pro: script beats and visual motifs. Use AI for shot ideas and alt lines. Keep your voice final.
- Production: mix human footage, practical elements, and AI passes. Record real-room audio to avoid sterile vibes.
- Post: color, pacing, typography, hands/faces/physics checks. Tag credits and provenance.
Move from outrage to output
AI isn't going away, and neither is slop. The fix isn't more rants-it's better work, shipped often, with clear ethics and visible craft.
If you want practical guidance on doing that, start here: AI for Creatives and Generative Video.
Your membership also unlocks:
