google's omni model + podcast ugc is actually blowing my mind this...

this wasn't edited together
i wrote one conversation, pressed generate, and got back a complete talking podcast in a single output
after dozens of tests i finally landed on the workflow that makes ai hosts sound surprisingly human
it's quietly become the highest converting creative format i've found, and i barely see anyone using it
so here's everything i use:
> my complete prompt from idea to finished video
> how i generate full 60 second podcasts in one shot
> the dialogue framework that keeps people watching
> the image setup that makes the avatars look real
rt + comment "podcast" and i'll send you everything
(follow for dm)