{
  "$type": "site.standard.document",
  "bskyPostRef": {
    "cid": "bafyreieqdt2whjh2sgaggjpyyshgsy6tocdixmgyxtaezyv63q6nnrn5ve",
    "uri": "at://did:plc:lk3jfj3zq4k4wxnk474axylu/app.bsky.feed.post/3mmld3xlbqb72"
  },
  "path": "/t/openai-is-making-a-huge-mistake-by-deprecating-dall-e-3/1367228?page=5#post_87",
  "publishedAt": "2026-05-24T04:42:16.000Z",
  "site": "https://community.openai.com",
  "textContent": "Let me just pop this in a carousel really quick and have a look how well it handles an edit command with this particular image…\n\nIt’s source fidelity is really low…\n\nI simply prompted to get rid of the deformity/lack of coherence…\nHaving it side by side will speed up my understanding slightly.\n\nAfter looking at that quick test I’m already convinced that before any serious attempt to replicate dall-e-3’s more difficult styles…\n\nRun the smaller dall-e-3 images through that sort of ‘upscaling’ with the current model first…\n\nIt already gives it more detail to work with / makes it easier for the model to understand the nuances in the image.\n\nI generally do the reverse engineering stuff with a set of three or more images so the model doesn’t tend towards trying to call out each individual entity but just from this test the style is handled already…\n\nIt just needs to be articulated.\n\nIt won’t likely ever get all the little fey creatures in the same places ever again but it can do this.",
  "title": "OpenAI is making a huge mistake by deprecating DALL-E-3"
}