Skip to main content
Participant
March 23, 2023

P: Show words omitted from prompt or violates content guidelines

  • March 23, 2023
  • 121 replies
  • 38714 views

Tried to make a word in colorful teddy plush but it says me it is against the guidelines. 

So I tried to say fur instead. That worked but didn't give me the result I wanted. 

121 replies

TomM1
Participating Frequently
April 15, 2023

Yes, boilerplate replies are annoying.

I wonder if Adobe is planning to evolve Firefly to the quality of some of the other AI sites or just keep it hobbled as it is.

Participant
April 15, 2023

Please stop repeating "Our goal is to provide a safe and enjoyable experience for all our users." It's neither helpful nor encouraging. You have comment after comment asking for Firefly to reply with what words are "bad" but no response as to whether you'll try to implement that. You have numerous requests for a list of words that are on the violation list, yet no concrete response. You're trying to provide a tool to artists, not children. For most artists, "safe" and "enjoyable" are probably at odds with one another.

_Waylo
Participant
April 14, 2023

when pressing 'Create' O get the following message

 

Uh oh

One or more words violate Firefly user guidelines and were removed.

 

It doesnt actually tell me what the words are. Looking at my prompt it looks safe enough. I'm trying to re-create my neighbour totoro for the DailyTheme -  "a 3 meter tall character shaped like a pear with grey dark fur and a white fur patch on its stomach and triangle shaped ears very small legs, stood at a bus stop in-front of some trees at night next to it on screen left is a girl holding a red umbrella, its dark and raining, standing next to each other, night, raining"

 

Would be good to show the words causing issue

 

Merging all similar threads here so the team can access all requests on this issue

João Pedro29400171excy
Participant
April 14, 2023

During the creation of prompts for creating images there are words that are interpreted as "sensitive", but there is no way for the user to see these words, to facilitate or replace that word with another.

 

I am a professional in the field of architecture, the first study I do is to gather some architectural styles that are not used together in a work, another principle that I use is to follow a logic "cool, the AI knows what a specific piece of furniture is, but it will be that she manages to apply a texture that is unnatural for this object", so you can see the ability of the AI.

 

Merging all similar threads here so the team can access all requests on this issue

 

Participant
April 11, 2023

"One or more words violates [....]" Never precise WHAT words are problematic. It is a real puzzle to try and find what words could be considered problematic, because most of them makes no sense. It would be very helpful to simply highlight problematic words in red. This way one can easily find them and find a synonym that isn't censored.


Thanks for reading.

Javad K
Participant
April 7, 2023

I wish it would specify the words that have problems
Or
delete it

 

 

Participant
April 5, 2023

The current prompting system is floored.
If there is a problem with a prompt (word or phrase) it needs to be highlighted so you know what to change.
There are also issues with names. Some (non English) names get flagged as language. Names should not be included in the language sector unless they are using non English characters.

Participant
April 3, 2023

Would be good to alert which words are causing Firefly to flag. For instance the following prompt contains "one or more words that violate Firefly user guidelines":

----

ancient castle, raging ocean, big waves, dramatic clouds, rough seas, waves crashing, cinematic, photorealistic

----

No problem with words being banned, but I also don't want to be banned as a user if I prompt the wrong words without knowing what they are. Thanks!

TomM1
Participating Frequently
March 31, 2023

Why doesn't Firefly like "Ominous" ?

 

Participant
March 31, 2023

I'd suggest displaying the word(s) that violate Firefly user guidelines in the "Uh oh.."-popup that shows so we can remove the word and stop the prompt from continously showing when generating multiple times using the same prompt