It depends. By default GPT-3 will skew towards generic outputs but you can mitigate that by giving it examples, more complex prompt structures, or chaining prompts together to replicate a creative process or framework. It doesn't replace human judgement, but in our experience with the right implementation the latest GPT models (3.5) can generate outputs that are on par with what a human expert might come up with.
Not right now but it will soon. We'll push an update on Substack when it is.
Hi Paul, Wondering if you made any progress on this generative approach to Brand Strategy?
What site is used for generating these pyramids?
Just a notebook and figma template right now but we will be launching a site that has this (and a bunch of other AI marketing tools) soon.
I think it’s awesome! Can it be tried somewhere?
It depends. By default GPT-3 will skew towards generic outputs but you can mitigate that by giving it examples, more complex prompt structures, or chaining prompts together to replicate a creative process or framework. It doesn't replace human judgement, but in our experience with the right implementation the latest GPT models (3.5) can generate outputs that are on par with what a human expert might come up with.