What music theory you actually need (and what you can skip)

Why this matters

You don't need a degree in music theory to make good AI music, but you do need vocabulary — the exact words that make prompts land. Genre names (shoegaze, phonk, boom bap, ambient techno). Structural terms (verse, chorus, bridge, drop, breakdown, coda). Instrument names (Rhodes, TR-808, koto, cello ensemble). Tempo ranges (75 BPM for a slow ballad, 174 BPM for drum and bass). Key relationships (A minor's a good default for melancholy, C major for anthemic). The rest — voice leading, secondary dominants, modal interchange — can wait. This task gives you the 40-word vocabulary that unlocks 80% of AI music prompt precision.

Demo

The 40-word music-theory-for-AI-prompting cheat sheet. Print it. Reference it every session for the next 30 days until it's second nature.

Try it yourself

1. Read the 40 words. Circle the 10 you don't know. Google definitions + a 30-second example clip for each.
2. Take one of your generated tracks. Describe it using ONLY words from the cheat sheet. What can you not describe? That's what's missing from the vocabulary.
3. Pick one genre from TEMPO_HINTS. Generate 3 tracks at wildly different BPMs (bottom, mid, top of range). Notice how BPM alone reshapes the vibe.
4. Watch one 5-minute music-theory YouTube video from Adam Neely or 12tone on a term from the sheet (e.g., 'syncopation'). Then apply it to your next prompt.

References

• undefined
• undefined
• undefined