SuicideFuel chat gpt's rating of my appearance

[deleted fren]

 

[Paladin]

From what I've seen chatgpt ratings can vary a lot depending on the prompt. I'm guessing it's programmed by default to tell you what you wanna hear

 

[deleted fren]

From what I've seen chatgpt ratings can vary a lot depending on the prompt. I'm guessing it's programmed by default to tell you what you wanna hear

that's true.

 

[Autist]

Fakecel

 

[AsakennyHao]

I'm too scared to ask anyone for rating

 

[deleted fren]

I'm too scared to ask anyone for rating

aren't you bald¿

 

[AtrociousCitizen]

What prompt do you use?

 

[zd60]

I'd be a bit careful with the output from chat gpt, it is very dependant on how you word the prompt, also good to test the same prompt on multiple LLM's to see if you get a somewhat consistent output

 

[AsakennyHao]

aren't you bald¿

No.

 

[lowz1r]

brutal because chatgpt usually gives inflated ratings even when you ask it to be honest.

 

[deleted fren]

What prompt do you use?

be brutally honest. Completely scientific. No sugarcoating or advice

 

[deleted fren]

I'd be a bit careful with the output from chat gpt, it is very dependant on how you word the prompt, also good to test the same prompt on multiple LLM's to see if you get a somewhat consistent output

good idea

 

[zd60]

good idea

That's what I do, testing multiple gives you a more refined output to draw your conclusion from, especially since chatgpt can sometimes be a bit sycophanty

 

[TooSomething]

I'm too scared to ask anyone for rating

Else they would expose you as a fakecel

 

[recessed]

just smile bro