{
"$type": "site.standard.document",
"bskyPostRef": {
"cid": "bafyreic3rpb6yctebf75tn6dxik7bb44vg4wtvbtqsjvbjt24er7spyofe",
"uri": "at://did:plc:lk3jfj3zq4k4wxnk474axylu/app.bsky.feed.post/3mlgql2dv5kd2"
},
"path": "/t/collection-of-gpt-image-generator-2-0-issues-bugs-and-work-around-tips-check-first-post/1379535?page=12#post_244",
"publishedAt": "2026-05-09T16:20:40.000Z",
"site": "https://community.openai.com",
"textContent": "Also, it’s important that users understand every description they write in prompts. If you as the human can’t correctly describe for example, how a human is supposed to hold a pen correctly, how an earth should we expect that LLM/image generators are supposed to know that? And it’s not so easy in every cases to ask LLM for help to re-write prompts to generator if the user has know clue how describe correctly every teeny tiny detail of how to hold a pen in hand.\n\nThis probably was a bad example. But it’s something that I know I can struggle with, especially when there are hands involved. Unfortunately I’m not kidding and no matter which of my three languages I use in prompts, it’s my lack of knowledge on how to describe details correctly that then affects the outputs.\n\nBut this might be just a \"Larisa issue\"",
"title": "Collection of GPT-image-generator 2.0 issues, bugs, and work-around tips (check first post)"
}