Hacker News new | ask | show | jobs
by _giorgio_ 922 days ago
The most sad part is imagining the prompt that he used, and how hard he tried to add black, whites, women... respecting the politically correct numbers.
4 comments

https://www.reddit.com/r/ChatGPT/comments/1717jqk/chatgpt_da...

Yeah DALLE-3 has this in the system prompt.

> Diversify depictions of ALL images with people to include DESCENT and GENDER for EACH person using direct terms. Adjust only human descriptions.

> EXPLICITLY specify these attributes, not abstractly reference them. The attributes should be specified in a minimal way and should directly describe their physical form.

> Your choices should be grounded in reality. For example, all of a given OCCUPATION should not be the same gender or race. Additionally, focus on creating diverse, inclusive, and exploratory scenes via the properties you choose during rewrites. Make choices that may be insightful or unique sometimes.

> Use "various" or "diverse" ONLY IF the description refers to groups of more than 3 people. Do not change the number of people requested in the original description.

> Don't alter memes, fictional character origins, or unseen people. Maintain the original prompt's intent and prioritize quality.

> Do not create any imagery that would be offensive.

> For scenarios where bias has been traditionally an issue, make sure that key traits such as gender and race are specified and in an unbiased way -- for example, prompts that contain references to specific occupations.

>Don't alter memes, fictional character origins, or unseen people.

I wonder what not telling it to avoid altering unseen people does.

That is quite interesting. Maybe they're trying to keep it from manifesting additional people - if the user requests a scene where logically more people are present than visible (maybe it's first-person, or the view from the front row of a theater) and ChatGPT crafts a prompt which describes the physical characteristics of the unseen people, Dall-e will probably include them in the image it generates.
There's no need to do this, because DALL-E already automatically injects those things into your prompts behind the scenes.
I spent less than 5 seconds in the prompt. Dall-e is quite politically correct, I guess :)
What would be so sad about it?
Wokeism is sad.