Apppublic

SeaWolf-AI/aether-ad-genesis

sourceHugging Faceapache-2.0updated 5mo agoView on Hugging Face

1likes

App README

AETHER-Ad Genesis v0.2

Emergent Advertising Idea Generation Engine integrating six theoretical pillars:

—Fauconnier & Turner — Conceptual Blending Theory (2002)
—Greimas — Semiotic Square (1966)
—Eberle — SCAMPER (1971)
—Stuhlfaut — Advertising Creativity 4D (2011)
—Boden — Surprise Dimension (2007)
—Coats — Pixar Story Spine (2011)

Part of the VIDRAFT / 지니젠AI AETHER meta-framework (emergence · metacognition · SLAI · collaboration · 상상상극).

Quick Start (local)

bash

pip install -e .
cp .env.example .env              # add HF_TOKEN or FIREWORKS_API_KEY
aether-ad generate \
  --product aether_ad/data/products/watch_a.json \
  --tension T01_power_asymmetry \
  --persona "10대 남학생, 좌절" \
  --duration 15 --n-seeds 30 --top-k 5 \
  --output outputs/watch_seeds.json

Space Demo

Select a product, tension archetype, and persona — the engine runs the 5-stage pipeline (Corpus → Encoding → Incubation → Emergence → Filtering) and returns the top-k scored ad-seeds as Pixar-spine storyboards.

Pipeline

ProductGenome  ─┐
                ├─▶ BlendingSpaces ─▶ CollisionRules(1–3) ─▶ PixarSpine ─▶ WowFilter ─▶ Top-k
ContextMatrix  ─┘         ▲                                                    │
                          └──────────── SLAI feedback (approved seeds) ────────┘

Architecture

See docs/ARCHITECTURE.md and docs/THEORETICAL_FOUNDATION.md.

License

Apache-2.0.