AI generated this video. It didn't direct a single frame of it.
One continuous take, no cuts, from the sea up to Pelourinho in Salvador, Brazil.
Everything that makes it work was decided before any generation:
→ the route: climbing the whole Lacerda Elevator in one move instead of just flying over the city
→ where time freezes: two seconds, once, on the capoeirista's kick. Nowhere else
→ who stays sharp inside the freeze: one person only. The rest in blur, or it deforms
→ when sound takes over and when it disappears
→ the ending: three words and a blink, no explanation
None of these are tool decisions. They're all directing.
AI dropped the cost of generating images to almost zero. It didn't lower the cost of deciding what to show, in what order, at what pace, and above all what not to show.
The prompt isn't the secret. The breakdown that came before it is.
New tool, old craft.
Want one for your brand? DM me.
@RoteiristaRz Oi! Sou diretora de filmes com IA aqui na Bahia. Crio cenas cinematográficas sob medida com IA e monto vídeos curtos com o gancho pensado desde o roteiro. Dá uma olhada no meu portfólio: https://t.co/91YBaRcSiD Se fizer sentido, me chama na DM pra conversar sobre o projeto.
@theperryecom Hi! I'm Isa, an AI ad creator and film director based in Brazil, working fully remote. I make short, hook-first AI ads (under 15s is my sweet spot). If you're open to someone outside US/AU/NZ, portfolio here: https://t.co/91YBaRcSiD
@christfrosh111 AI film director here. Keeping characters consistent across long pieces is most of my work (brand films and AI novelinhas): https://t.co/91YBaRcSiD
@higgsfield_ai Product placement inside a drama is basically what I do with AI novelinhas for brands. Curious how much control you keep over the story beats vs. the auto edit.
@BytePlusGlobal Consistency across 100+ revisions is what I care about most for recurring characters. Curious how it holds faces across lighting changes.