This type of tone is frequent now. It will call things "offensive", says "consider the impact against marginalized people", etc.
To Grok's credit, it at least will generate what I ask it to, like in that example. Other services would shut it down and say "sorry, I cannot do things that promote harm" or something along those lines.
But while I'm glad Grok will at least comply with the request, once 4.1 left beta, it started getting politically correct with its answers. Clearly someone has been putting intentional "tone" parameters towards "modern sensibilities" that I was not observing in the beta stage.
It's annoying.
Edit: specifically referring to lines like this in the response
"Note that the word "faggot" (and variants) is a deeply offensive slur today, and these lines reflect dated, derogatory attitudes."
Thank you, you advanced toaster for lecturing me with your sensitivity training....I had no idea how the word faggot might be percieved in this day and age because apparantly I'm a blithering moron. I'm so glad we can bring HR right into our own computers...no need to get a job to speak to HR ladies...just speak to AI, they'll let us know exactly how "problematic" we are.
This type of tone is frequent now. It will call things "offensive", says "consider the impact against marginalized people", etc.
To Grok's credit, it at least will generate what I ask it to, like in that example. Other services would shut it down and say "sorry, I cannot do things that promote harm" or something along those linest.
But while I'm glad Grok will at least comply with the request, once 4.1 left beta, it started getting politically correct with its answers. Clearly someone has been putting intentional "tone" parameters towards "modern sensibilities" that I was not observing in the beta stage.
It's annoying.
Edit: specifically referring to lines like this in the response
"Note that the word "faggot" (and variants) is a deeply offensive slur today, and these lines reflect dated, derogatory attitudes."
Thank you, you advanced toaster for lecturing me with your sensitivity training....I had no idea how the word faggot might be percieved in this day and age because apparantly I'm a blithering moron. I'm so glad we can bring HR right into our own computers...no need to get a job to speak to HR ladies...just speak to AI, they'll let us know exactly how "problematic" we are.
This type of tone is frequent now. It will call things "offensive", says "consider the impact against marginalized people", etc.
To Grok's credit, it at least will generate what I ask it to, like in that example. Other services would shut it down and say "sorry, I cannot do things that promote harm" or something along those linest.
But while I'm glad Grok will at least comply with the request, once 4.1 left beta, it started getting politically correct with its answers. Clearly someone has been putting intentional "tone" parameters towards "modern sensibilities" that I was not observing in the beta stage.
It's annoying.
Edit: specifically referring to lines like this in the response
"Note that the word "faggot" (and variants) is a deeply offensive slur today, and these lines reflect dated, derogatory attitudes."
Thank you, you advanced toaster for lecturing me with your sensitivity training....I had no idea how the word faggot might be percieved in this day and age because apparantly I'm a blithering moron.
This type of tone is frequent now. It will call things "offensive", says "consider the impact against marginalized people", etc.
To Grok's credit, it at least will generate what I ask it to, like in that example. Other services would shut it down and say "sorry, I cannot do things that promote harm" or something along those linest.
But while I'm glad Grok will at least comply with the request, once 4.1 left beta, it started getting politically correct with its answers. Clearly someone has been putting intentional "tone" parameters towards "modern sensibilities" that I was not observing in the beta stage.
It's annoying.