Skip to main content
Participating Frequently
March 24, 2024
Open for Voting

Stop Being so Woke

  • March 24, 2024
  • 1 reply
  • 481 views

This is just like Google's AI image generation where it forbid the imaging of white people. I typed in President George Washington and it gave me two images of women and two images of non-white men.

Freaking disgusting!

This is a terrifying form of thought policing.

1 reply

rayek.elfin
Legend
March 25, 2024

It has nothing to do with that non-sensical word that seems to change meaning whenever the person uttering it feels like it proves something they disagree with.

 

Simply stated, Adobe doesn't train its genAI model with faces from well-known / famous people to avoid abuse and potential litigation or bad rep in the media. So the words "George Washington" do not mean anything to its AI model and it will produce random output.

 

Just imagine if someone would create an image of George Washington behaving in a vile way or it is abused as a tool for political fake messaging. Then Firefly and Adobe would become a target if it is discovered Firefly was used to create that image. That is bad for business.

 

That's all. I am not saying I agree or disagree with this. But from a business point of view of a large company like Adobe: yes, it does make sense to de-weaponize potential abuse. Pink bunnies holding hearts? Sure! Sun-set landscapes with smiling people? Sure! George Washington pointing a gun at another famous person? Nope. No can do.

And that, surely, is totally understandable.

 

If you do not like this, install open source tools on your local machine or use different tools. That is your prerogative. You have that freedom.