01 / Challenge
Challenge
The lazy speaker reuses one deck for every audience. But teachers, enterprises, governments, and founders speak different languages, carry different pains, and respond to different examples. Preparing four high-quality talks in four days is humanly impossible without AI — but using AI without customization yields hollow, generic content. The real question: can AI-produced content actually enter the audience's mind?
02 / Approach
Approach
Audience-profile research first (Gemini/Claude) to surface the real pains and language of each group. Core framing stays constant; the AI rebuilds opening hooks, examples, dialogue exercises, and CTAs per audience. Each session has a clear engagement angle — for teachers: student engagement; for enterprises: ROI; for government: public value; for founders: speed advantage. After each talk, immediate post-mortem feeds the next.
03 / Result
Result
Across four sessions, a consistent reaction: "this talk feels like it was made for us." Same core message, four different packaging strategies, four equally strong resonances. AI shifted from "content generation tool" to "audience insight + customization engine."
Why it matters
Why it matters
"Using AI to make slides" is Basic. "Using AI to make good slides" is Proficient. **"Using AI to craft slides that move each audience" is Advanced** — the metric shifts from "did I produce" to "did the audience absorb." OUTCOME = audience resonance; IMPACT = knowledge actually entering different communities. This is Use · Advanced: integrating the audience's real pain.