A speaker explains creating custom GPTs for clients by loading brand information (tone, voice, avatars, strategy) to generate tailored marketing content (e.g., podcast intros). The process involves iterative prompting, training, and review to save time and shift lower-level tasks away while making higher-level roles more strategic.
We use AI, but we use it from the prospect of saving time, right? So we do use ChatGPT, but we have this workaround where we create what is called a GPT for our client and we put all of our brand information in there, right? And we say, this is the tone, this is their voice, these are their avatars, this is their strategy, what they're promoting, this is how they speak. So we promise prompt the hell out of it. And then we have this trained GPT that we could work with and say, hey, Flip and Danny are releasing a new podcast. And based off what I've trained you on, the podcast description, please add in these elements. Q Muldrick is the speaker. This is her bio. Here's the transcript. Create a podcast intro, but that required me going in and training it and prompting it and reviewing it. So what it's doing is it's hedging out lower level jobs. I will say that, but it's making these higher level jobs be more strategic.
Physicians and High Earners — What Real Estate Can and Can't Do for Your Taxes
