A Wizard of Oz landing

Carlos, the socio, has a new idea: offer companies servers to run their own local AIs ▸ 20:38. It’s plausible, capital-intensive, and completely unvalidated, the exact profile of idea that eats savings. Juan’s prescription comes with its name attached: “hágase un mago de Oz… cuando uno no tiene nada hecho, pero publica algo, como pescando” ▸ 20:49. A landing page for the product that doesn’t exist, with a curtain where the machinery should be.

The full circuit, as prescribed ▸ 21:16: build the landing, put a little advertising behind it, watch whether anyone registers or asks for information. If real interest shows up, then compute costs, send an actual quote, and only then worry about capital, “si necesitamos plata, le pedimos al padrino” ▸ 21:37. Every expensive step is gated behind a cheap signal from the market.

la cortina primero, la maquinaria después →

Note the delegation too. Juan hands Carlos the whole play and deliberately doesn’t take the project on, “yo por ahorita no quiero empezar otro proyecto”, offering to help with copy and ads only “si lo veo animado” ▸ 21:54. The smoke test validates the idea; whether Carlos builds the landing validates Carlos. Both experiments run on the same page, and neither costs Divo’s launch week a single hour…