When signing up I got a warning that this platform can’t be used to talk about ChatGPT and I respect that not asking why but how come there are so many posts about ChatGPT site/app and nothing is being done about it? Seems very inconsistent moderation first you disallow community to talk about your product then still allow it lol
Welcome to the forum, @Slavax …
This post gives a lot of history about this place if you have the time.
Weclome to the forum!
You are absolutely right to make that observation and ask the question.
While there are likely several reasons for this, one I have noticed is that the forum sometimes allows topics and categories to evolve based on what the community finds useful, rather than adhering strictly to the stated purpose you quoted.
Okay thank you very much guys I just didn’t want to break any rules @PaulBellow @EricGT
They have blended a lot of products now. ChatGPT, OpenAI Codex, etc. So separation of tools is harder in 2026.
Yep that note needs an update since the products have been merged somewhat.
It’s still a developer-focused forum though!
By intent yes, by usage no.
At first impression as anonymous user I might think I have landed in the complaint dept. I note anonymous user arriving via a Google search or something because they would not see the noted message in the first post.
By tags, and meaningful questions in the recent weeks has shifted more to Codex.
By likes, the image gallery is winning by far.
I come here for the art and the experimental atmosphere, my favorite users that I read most are @sergeliatko @mitchell_d00 @phyde1001 @Tina_ChiKa @DysTopia @LarisaHaster and @jochenschultz, but everyone does a great job here!
Here are some developer/image related tags
[projects](/tag/projects)
[project](/tag/project)
[Codex](/tag/project)
[image-generation](/tag/image-generation)
No, he is saying most users just come here from Google and lodge a complaint. Or come from Google and only see folks lodging complaints. New users don’t understand tags. They don’t see start pages coming from Google.
And he is doing user demographics by likes and tags. I do that too lol.
You can “feel the water”
Just as an example of why this isn’t really a “ChatGPT discussion” — but very easily could be mistaken for one…
The forum has a long-running topic built around prompts such as:
The interesting part isn’t the ChatGPT website.
It’s the experiment: what has the model inferred about the person, and how does it turn that accumulated context into something new?
You can perform essentially the same experiment with music.
I asked:
“Find me some music on YouTube I might like.”
It returned:
Radiohead — Weird Fishes / Arpeggi (From the Basement)
Massive Attack — Teardrop
The Chemical Brothers — The Test
Woodkid — Run Boy Run
Nick Cave & The Bad Seeds — Into My Arms
GUNSHIP — Tech Noir
(Let’s be clear I haven’t even reviewed them yet this is a blind test)
Then:
“Meaning of each of those songs”
And now there is an actual experiment to discuss.
Why those six?
Was it matching musical taste? Things I’ve written about? Recurring ideas? Family, technology, protection, change, nostalgia, determination? How much came from context, and how much was just plausible model association?
That’s analogous to:
“Based on what you know about me, draw a picture…”
One asks the model to represent its impression of you using pixels.
The other asks it to express its impression of you through music recommendations.
And this is also why I wouldn’t call the experiment itself “ChatGPT.”
ChatGPT happens to be the interface I used to run it.
Otherwise we’d all just be copying the same implementation instead of experimenting with the underlying idea.
The underlying thing worth discussing is the idea, behaviour, model, context, algorithm or experiment — not necessarily the particular application window through which somebody happened to access it.
That’s where I think the boundary gets interesting:
“How do I use feature X in the ChatGPT app?” → ChatGPT product discussion.
“What happens when an AI uses accumulated context to infer my tastes and represent me through images/music/code/etc.?” → much broader AI experimentation.
Same interface.
Very different subject.
This post was flagged… Interesting…
For an example beyond the coding/developers of “Codex”, we have where “OpenAI Platform” is offered as a new choice in ChatGPT. This is an example of “merged platforms” - but one that makes no sense, especially considering the possible hierarchy of API organizations vs ChatGPT accounts, and what you do NOT want ingested as training data.
It is surfaced in ChatGPT where you have a pop-up to add a file or add your personal data OpenAI wants to dip their toes in.
When it does not serve an API developer in the slightest, or the AI model of ChatGPT only produces a preamble and does nothing in conjuction with this connector, you can report it here.
Below, blue messages are user prompts to try to get the AI to produce something describing what and why this “platform” is in ChatGPT.
(had to switch to “pro” instead of “xhigh” that immediately stopped instead of going agentic).
Thus, I might get flagged for showing examples of ChatGPT’s output, here I report on something offered specifically for API developers (that also needlessly subscribe to ChatGPT?).
Oh wow! This really highlights how blurry the boundaries become when ChatGPT accounts and OpenAI Platform/API organizations are merged into the same interface. If the connector doesn’t actually provide useful access to Platform resources, it creates more confusion than functionality. At minimum, OpenAI needs to make the permissions, organization hierarchy, and treatment of connected data crystal clear. Can of worms..
I’m out of hearts @phyde1001 ill be back, yum custard!
Thanks for pointing out the confusing wording. We’ll pass the signup-message feedback to the community team so they can clarify the forum’s scope. We don’t have a timeline to share.