Week 1 Semester Kickoff

TL; DR

• Summer website model (aidolworld.ai) bake-offs done
• In addition to technology and narrative design pillar, we decided to add an economy as a third pillar.
• Engine pinned: Unity 6.0 (6000.0.82f1)

Please read project overview here before continuing.

· · ·

We didn’t start from zero.

Over the summer we built a web prototype at aidolworld.ai — the idol-training loop in its simplest possible form, running on real APIs end to end.

Interview -> Image Generation -> Music Production -> Debut Performance Stage

Then we handed it to twenty or thirty friends and classmates, and watched.

More than half of them asked if they could download their idol’s images and music.

That’s the finding.

There was barely any narrative in the summer build. No story. No stakes. No four-week arc. No economy. Just a loop and some generation.

And over half the people who played it wanted to keep what came out.

Players are possessive of their virtual assets and their virtual identity. The attachment forms far earlier, and far more cheaply, than we expected — and once it exists, people want to hold the artifact in their hands.

What we changed because of it:

If attachment is the thing that reliably happens, the design job is to build for ownership.

So, we came back with a third pillar:

Pillar	Status
Narrative design	Continuing from spring
AI pipeline	Continuing from spring
Economy	New this semester

We don’t only want to document how a small team builds an AI-driven game. We want to document whether it can pay for itself.

· · ·

About the money

Everything above was self-funded.

The summer prototype, every bake-off, every exploratory generation, out of Cheryl’ own pockets, before the semester started.

This semester we’ve requested $7,000+ from the CMU Entertainment Technology Center. It’s submitted and in review.

	Amount
Part 1 — Team Efficiency Subscriptions	~$3,917
Part 2 — API/playtest caps	~$3,090–$3,650
Total ask	~$7,007–$7,567

· · ·

Weekly Progress Report

Meeting Notes:

On the game

• Three systems are still undesigned: character attributes, fan growth, lucky events.
• Balance skill against luck. Players need their own effort and lucky moments.
• Quantify the pleasure. He wants a table: which mechanics and which visuals actually produce a rush.

On the road ahead:

• Opening a conversation with Suno about a possible collaboration
• Start preparing for SXSW
• Start early on the continuation pitch — progress is what determines approval