{
  "$type": "site.standard.document",
  "bskyPostRef": {
    "cid": "bafyreieg4xjz3pitjejra6rndwate3yl73ppxydxhyu6a5hrbgcpdc666y",
    "uri": "at://did:plc:lk3jfj3zq4k4wxnk474axylu/app.bsky.feed.post/3mlw4fkh5bsw2"
  },
  "path": "/t/longer-context-windows-gpt-for-creatives-and-better-project-integration/1381021#post_1",
  "publishedAt": "2026-05-15T19:02:00.000Z",
  "site": "https://community.openai.com",
  "textContent": "I mostly use ChatGPT for creative writing, worldbuilding, long-form personal writing, and ongoing canon development. I am not an API power user. I signed up here because this seems like one of the only places where OpenAI actually takes structured feedback seriously, even though most of my daily use is in the ChatGPT app.\n\nThe issue is simple: ChatGPT is increasingly optimized for enterprise, coding, and developer workflows, while long-form creative users are stuck fighting the product instead of using it.\n\nThe biggest problem is the conversation limit.\n\nFor creative writing, worldbuilding, personal journaling, long-form planning, or ongoing fictional universes, the “you’ve reached the maximum length for this conversation” wall is incredibly frustrating. It breaks the flow, forces a new thread, and destroys continuity. When you are building a fictional canon, writing a long emotional arc, or developing a story universe over weeks or months, you do not want to keep starting over.\n\nI understand there are technical reasons for context limits. I understand models can degrade over very long conversations. But for many creative users, we are not asking the model to keep 7,000 lines of code perfectly indexed forever. We often just need basic continuity: who the characters are, what the established canon is, what relationships exist, what has already happened, what tone we are writing in, and what the active thread is building toward.\n\nRight now, that experience is fragmented.\n\nProjects should solve this, but they currently feel more like storage folders than actual creative workspaces. A project can hold files and chats, but the chats do not meaningfully talk to each other. Starting a new thread inside a project often feels like starting from scratch. If a conversation hits the limit, you cannot simply keep rolling inside the same project with the existing context intact. That makes Projects feel borderline useless for large creative work.\n\nFor example, if I have an alternate-history aviation canon, I should not need five or ten separate conversations to keep developing it. If I have a long-running personal writing project, I should not need fifteen different threads for the same emotional arc. I should be able to keep one rolling project thread, or at least have new threads inside that project inherit the project’s active canon, summaries, character relationships, terminology, and previous decisions.\n\nWhat I would love to see is a “GPT for Creatives” or “Creator Mode” subscription/workspace.\n\nNot necessarily for coding. Not necessarily for enterprise. For writers, worldbuilders, roleplayers, essayists, podcasters, journaling users, lore-heavy creators, and people using ChatGPT as a long-form creative partner.\n\nCore features could include:\n\n  1. Rolling creative context\n\nA way to keep a project going without constantly hitting a hard wall and being forced to restart. Even if the full conversation cannot remain in live context forever, the project should preserve a reliable working memory of the canon.\n\n  2. Project-level memory\n\nEach project should have its own persistent memory separate from general ChatGPT memory. A project should remember characters, timelines, world rules, tone, terminology, previous plot beats, and user preferences within that project.\n\n  3. Canon/source-of-truth files\n\nLet users upload or write a “project bible” that the model treats as authoritative unless explicitly overridden. For fiction, this could include timelines, character sheets, factions, tech specs, locations, relationship maps, and style rules.\n\n  4. Better cross-thread continuity\n\nIf I start a new chat inside a project, it should know what the project is about. It should not behave like the prior chats are sealed off from each other. The whole point of a project should be continuity.\n\n  5. Conversation continuation after limit\n\nWhen a thread hits the maximum length, there should be a “continue this conversation in same project” button that automatically carries forward a high-quality summary, key decisions, active characters, open loops, tone, and source files.\n\n  6. Creator-oriented organization\n\nGive us timeline views, character indexes, lore folders, summaries, tags, pinned canon notes, and the ability to mark something as “canon,” “draft,” “alternate version,” or “discarded.”\n\n  7. Better memory control\n\nCreative users need to be able to say: “Remember this only for this project,” “Make this canon,” “Forget this version,” or “Use this as the current continuity.”\n\n  8. Long-form personal writing support\n\nThis is not only about fiction. Some users use ChatGPT for long-form venting, journaling, emotional processing, memoir-style writing, or ongoing personal reflection. Those workflows also suffer when every long conversation eventually gets cut off and scattered across multiple disconnected threads.\n\n\n\n\nThe frustrating part is that ChatGPT is already close to being incredible for this. It can remember tone. It can track characters. It can develop scenes. It can help organize huge creative ideas. But the product structure keeps interrupting the very workflows where ChatGPT feels most useful.\n\nRight now, OpenAI seems heavily focused on enterprise and coders. That makes sense as a business priority, but it leaves a huge user base underserved: people who use ChatGPT as a creative partner.\n\nMany of us are paying subscribers. We are not trying to abuse the system. We are not asking for infinite compute for no reason. We are asking for a product tier or feature set that recognizes that creative work is long, messy, iterative, emotional, and continuity-heavy.\n\nA real creator-focused ChatGPT workspace would be a game-changer.\n\nProjects should not just be folders. They should be living creative workspaces.\n\nA project should know its own canon. It should preserve continuity. It should let long-running work keep going without forcing users to constantly rebuild context from scraps.\n\nThat is the feature I would pay for.",
  "title": "Longer context windows, GPT for creatives, and better project integration"
}