ZJUAANC ZJUAANC

VI AI Music Academy Lecture 2: Structure and Emotion — Chorus, Conflict, and Holding 20 Seconds (plus Jev; guest Fran)

Lectures · AI Music Academy

Joint community event by ZJUAANC and VI AI Music Academy.

Event Details

  • When (Three Time Zones):
    • PDT: Sat 26 Sep 2026, 5:30–6:30 PM
    • EDT: Sat 26 Sep 2026, 8:30–9:30 PM
    • Beijing Time: Sun 27 Sep 2026, 8:30–9:30 AM
  • Where: Zoom Online Meeting (Zoom link is in the Luma confirmation)
  • Nature: Free Community Event. Zoom. 60 minutes, including a 5-minute break.
  • RSVP (This Lecture): https://luma.com/z0329spm
  • Academy Website: https://music.weiai.ai

Free. Zoom. 60 minutes, including a 5-minute break.

Lecture 2 of 5 in VI AI Music Academy Season 1.
Lecture 1 asked: is this draft editable?
Lecture 2 asks: will a listener stay for 20 seconds — chorus, conflict, hook.

Guest: Fran, an award winner in a China AI music creation contest, will speak briefly on what the process felt like and what she would tell someone still revising a first batch of songs.

When

  • PDT: Sat 26 Sep 2026, 5:30–6:30 PM
  • EDT: Sat 26 Sep 2026, 8:30–9:30 PM
  • Beijing: Sun 27 Sep 2026, 8:30–9:30 AM

RSVP (This Lecture)

https://luma.com/z0329spm
Zoom is in the Luma confirmation.

Lecture 1 (ended): https://luma.com/d96kvs6t

Season 1 Curriculum

  1. Sep 12: From Idea to an Editable Track: Structure, Lyrics, Style Tokens, Negative Prompts (Ended)
  2. Sep 26: Structure and Emotion: Chorus, Conflict, and Holding 20 Seconds (This Lecture)
  3. Oct 10: Chinese Vocals and Traditional Instrument Timbres
  4. Oct 24: Arrangement Layers and Cinematic Imagery
  5. Nov 7: Final Delivery: Version Selection and Documenting Revisions

Clock

  • 5:30–5:50 Chorus function, conflict curve, 20-second hook. Compare two cuts. Pass/fail on “does it hold,” not taste.
  • 5:50–6:00 Guest Fran: contest experience, what she noticed later about structure and emotion, practical advice (8–10 minutes)
  • 6:00–6:05 Break
  • 6:05–6:25 Agents help rewrite structure; they do not ghostwrite the song. Brief on Jev (TypeSafe AI, 15 Sep 2026): a System One decision model. It returns typed scores, not lyrics. Generation stays with an LLM / Suno-class tool; judging a hook can be a Jev-class call.
  • 6:25–6:30 Q&A

What You Will Take Away

  • A 20-second checklist: entry, persona/setting, conflict, chorus promise.
  • Three reusable prompts for structure rewriting rather than asking for more adjectives.
  • Jev’s role in the creative loop: LLM generates candidates, decision model scores. It cannot hallucinate out-of-schema text, but human judgment verifies the final cut.
  • Fran’s field insights: which section to rework, what to listen for, and avoiding the trap of one-shot perfection.

Guest Speaker: Fran

Award-winning creator in a domestic AI music creation competition. In this lecture, Fran will share three things: her authentic experience going through the competition, what she later realized about structure and emotion, and practical advice for creators revising their first batch of tracks. No contest entry prerequisites, no scoring rubrics.

Jev in One Paragraph (For Non-Technical Audiences)

Released by TypeSafe AI on Sep 15, 2026 (current public release approx. jev-1.13). Jev is not a chatbot. You send state plus typed questions; it returns a choice, a score, or a yes/no with confidence, in roughly 70–500 ms. It is billed per input. It fits narrow assessments like “does this chorus hold a hook?” rather than “write this song for me.” This session covers its practical boundaries without sales pitches or registration requirements.

Differences from Lecture 1

  • Lecture 1: What makes a draft editable, and general RSI (bounded self-improvement exists; unconstrained models designing next-gen models has not occurred).
  • Lecture 2: Audible standards for structure and emotion, decoupling generation models from decision models, and a guest reflection from an award-winning creator.

Speaker: Michael Huo
Chair of VI AI Foundation & VI AI for Good Foundation; Founding Board Member & Vice President of ZJUAANC; Chair of NACUAA AI Venture and Impact Club.

Guest: Fran, Award-winning AI Music Creator

Organizers: Zhejiang University Alumni Association of North Carolina (ZJUAANC), NACUAA AI Venture and Impact Club
Executive Organizers: VI AI Foundation, VI AI for Good Foundation, AI Agent Alliance
Co-Organizers: NACUAA North Carolina, South Carolina, Silicon Valley, Northern California, DC, and Washington State Chapters; Sister ZJU Alumni Associations of Tennessee, Philadelphia, Rhode Island, and Florida
Media Partner: North Alpha Post