Totally agree, ChatGPT's default friendliness can be a double-edged sword. It’s great for tone but often gets in the way of honest critique or pushing our thinking. This post gave me some solid ideas to test for sharper AI output.
This is super helpful, but quick clarification--above, it says to paste this at the start of every prompt. Should we do this AND customize our Traits using the guidance from your other article? Or should we just use this prompt and not worry about Traits?
You might consider using this as prompt instructions in a Project / Gem or as the instructions for a custom GPT (in ChatGPT), so you don't have to copy/paste this prompt in every time.
I find this and a few others like it to focus too much on being "ruthless" and I struggle to get useful feedback / output as a result. It's like asking an asshole co-worker when I want the honest but helpful and constructive one.
That said, this is a solid technique overall and the smart marketer will take this into their AI of choice to tweak and improve it for their own needs (as I did). So, thanks!
This is gold. Thank you.
Glad it helped!
Totally agree, ChatGPT's default friendliness can be a double-edged sword. It’s great for tone but often gets in the way of honest critique or pushing our thinking. This post gave me some solid ideas to test for sharper AI output.
100%. I realized friendliness ≠usefulness. Sometimes we need a tool that pushes back & not pats us on the back.
This is super helpful, but quick clarification--above, it says to paste this at the start of every prompt. Should we do this AND customize our Traits using the guidance from your other article? Or should we just use this prompt and not worry about Traits?
I would say do both!
Traits = setup once
This Prompt = Do it once at the beginning for every new chat
Thanks!
You might consider using this as prompt instructions in a Project / Gem or as the instructions for a custom GPT (in ChatGPT), so you don't have to copy/paste this prompt in every time.
I find this and a few others like it to focus too much on being "ruthless" and I struggle to get useful feedback / output as a result. It's like asking an asshole co-worker when I want the honest but helpful and constructive one.
That said, this is a solid technique overall and the smart marketer will take this into their AI of choice to tweak and improve it for their own needs (as I did). So, thanks!