also given enough fine tuning and effort, you could probably make it work with babbage or curie, further reducing the cost. I’m sure they have their reasons for using a wordlist but having somthing that can understand context would help a ton. Maybe use a wordlist as a backup gpt fails? The added latency shouldnt be a problem since it already takes a bit of time to generate the image.
1 Like
I highly doubt that. It’s simply easier to implement (no train - just yeet wordlist in and done) and costs nothing, but that’s not a valid reason imo.
Meanwhile tens of people have already been banned for no reason at all, hurting OpenAI’s revenue. Most don’t come back after that, considering that there are alternatives which are cheaper, work almost as good and the user doesn’t need to wait a team’s manual decision to get unbanned.
It’s kinda sad. I love OpenAI, but this decision in particular I just don’t get it.
2 Likes
I had my account suspended, filed an appeal a week ago. Was just trying to edit the rib of an angel with inpainting. I can’t figure out anything wrong I’ve done.
4 Likes
I got revoked for a description of chicken in sauce which I had Chatgpt write… which part of this triggered the filter? Chatgpt violated policies of Dall-e.
“A sleek, modern restaurant plate is set with a beautifully plated portion of chicken in creamy white sauce. The chicken is tender and juicy, with a golden, crispy skin. The sauce is a pale, creamy white color, with swirls of dark, earthy spices throughout. A bed of creamy mashed potatoes is spooned onto the plate, and a sprig of fresh herbs adds a pop of vibrant green color. The dish is garnished with a sprinkle of crispy fried onions, adding a crunchy texture to the dish. The plate is illuminated by a dim, atmospheric light, and a glass of chilled white wine sits nearby, ready to be sipped alongside the dish.”
would be my guess… in context it’s fine, but out of context… well, I’ll leave that to your imagination. Haha.
You might try a shorter prompt, too.
Good luck!
1 Like
I’m getting these safety errors for recipe prompts generated by davinci… not sure how to get around it, here is the text (French) nothing fancy…
Recette de Spaghetti aux Tomates Cerises, Poulet et Parmesan
Ingrédients
- 500g de spaghetti
- 500g de tomates cerises
- 200g de parmesan râpé
- 2 blancs de poulet
- 2 gousses d’ail
- 2 cuillères à soupe d’huile d’olive
- Sel et poivre
Instructions :
-
Faites cuire les spaghetti selon les instructions sur l’emballage.
-
Pendant ce temps, faites chauffer l’huile d’olive dans une grande poêle et faites-y revenir les blancs de poulet coupés en cubes jusqu’à ce qu’ils soient bien dorés.
-
Ajoutez les tomates cerises coupées en deux, l’ail haché et le sel et poivre. Faites cuire pendant 5 minutes.
-
Ajoutez les spaghetti cuits et mélangez bien.
-
Ajoutez le parmesan râpé et mélangez bien.
-
Servez chaud et dégustez !
“The Netherlands is known for its dikes”. Is problematic for DALL-E 2. 
Replaced dikes by dunes, and it worked. So, “dikes” are the problem.
Extra test: The Netherlands is known for its dunes and dikes. Indeed, not working.
Get the filters correct!
2 Likes
It seems ChatGPT violated the policies of some other competitors as well 
Further testing proved that the issue is with the word “skin”, which I replaced with “texture”, and it didn’t throw any policy violation errors.
1 Like
This prompt used to work “Vestal virgins shopping for clocks at a flea market.” on Dall-e Mini. But I just tried it and it was rejected by the safety system, I presume because of the “virgin”, which is kind of … Victorian.
1 Like
I know this is an old thread but I ran into the same problem a while back when I tried giving it my own long-winded description of a New Yorker cartoon, a play on the painting Nighthawks by Edward Hopper. No matter how I rephrased it, I got a policy violation warning, and when I insisted, the threat of suspension, so I just stopped, better safe than sorry. (I’d have loved to see how DALL-E would interpret my words, recreating – or not – the cartoon.)
1 Like
Is there any indication that the content policy will be updated, it currently doesn’t explain what’s wrong with “a portrait painting of elon musk with a clear plain white background”, when it works for “a portrait painting of morgan freeman with a clear plain white background” and also works for “a portrait painting of mel gibson with a clear plain white background”.
1 Like
Completely concur that the content filters are bu//$H!73. Paying member, banned from Dall-E for months now due to not having clear guidelines.
Very frustrating.
2 Likes
I have yet to read about anyone winning an appeal - Did you?
1 Like