Allow users publish AI Bot PM as new topic

After PMs with AI there can — and is for me — need to convert PM topic to public topic. When an user discuss with AI, and that topic is worth to show public, there is few tools to achieve that:

  • flag and ask admin to do it
  • invite admin to that topic
  • copy&paste and posting it as own topic

First and second are possible solutions, but those craves too much manual labour in such trivial task. And are behind too many steps for ordinary users.

The third one in not an option, or it must be the very last option. I dislike very deeply nowaday’s style to claim a text made by author itself when everybody see right away it is AI-generated [1].

More honest solution would be ability convert that PM-discussion to a public topic.

There can be some caveats. Like very basics of Discpurse doesn’t allow such action easily or at all. Similar situation than with several drafts, that can’t be fixed and that’s why we have only bad or worse options. And if that is the situation then let it be.

There can be some juridical questinmarks. Up here in northern Europe every private persons can publish private conversations no matter if it has happened via snailmail, phone, email, SMS, PM… media doesn’t play any roll whatsoever. That doesn’t mean everyone can publish what ever they want, but that would be weighted after, not before. But Europe, or even one part of it, is not same as the world and there can, and propably is, local restrictions.

There can be some moral issues, though. And I know it is really american way to limit user’s action and content because of that, but that should be headache in the States only.

But AI is not a person. But I can just imagine how difficult would be allow publishing PM with AI and at same time not allow it with a real user. Can we use AI to solve out that dilemma :smirk:

I have bad feeling this is just another version of drafts where we should copypaste PM to topics, that has its own limitations.

So. If converting a PM to a public topic is technically possible, could it happend at some point?

Or if not I’m open to any suggestion how to make that operation as smooth to an ordinary user as possible.

  1. yes, I know out there is a lot of admins and and content creators who thinks that way is wanted, but, sorry to say this at loud, it is just strong flag showing lazyness and williness to do everything just for ads; no breaking news here, because internet is full of copied content with same errors and typos ↩︎


Yeah. I noticed they need too. I was thinking of making a plugin that would let a regular user convert a pm to topic if the only other participant was an AI. (And I still might if I do a project that’s probably going to happen)


There might be an interesting idea.

There are 2 plugins that might be able to be used. There is a plugin that allows Op of a topic to have some moderator powers over their own topic; maybe a feature request or use the Private Topics plugin might be better

This one above might be easy as I presume you might still be able to edit category to a public one


Hello :wave: I was thinking about this too recently. It seems that we are on the same wavelength of AI testing :sweat_smile: In my case, I came up with a slightly workaround solution, if you can call it a solution. Since we assume that personal communication with AI may someday become public, then the Forum Administrator himself can start a private conversation (as I did) with AI, and then add a certain group of people to the conversation, having previously notified them that this conversation may become (possibly will definitely become) publicly available in the future. And to maintain relative confidentiality (if you can call it that) in some topics I use (still experimentally) hidden replies plugin + I use hiding tags on a message to hide some attributes (including usernames). Thus, after publishing a topic publicly, other users will only be able to see the AI’s answers (if it is an artist, then users will be able to see an additional expandable statement according to which the answer was given).

P/s. In the meantime, let me suggest you read this topic which I recently published , maybe she will be interesting to you :slightly_smiling_face:


I am not sure about this, particularly, when sharing AI interactions you often want to share a chain:

  • User asked AI: X
  • AI answered: Y
  • User asked for clarification: X
  • AI answered: Y

Then sharing that entire exchange in one post is really what a lot of people desire, maybe even inside a [details]

I wonder @tynaut / @Falco / @Saif if an interesting option here is an extra button here:


You click it to share, and it would allow you to optionally pick how much context you want to include.

Then we can even have some special styling for these snippets of AI based conversation.


People are… not lazy, but ordinary johns and janes want as easy and streamlined path as possible. For admins, moderators, actives and IT-oriented such choose-edit-customize operations aren’t any bigger thing.

And actually whole conversation can and will be very valuable quite often. What is this → AI-answer → what if that → AI-answer → tell me more → AI-answer → forum guys, is this, that, those…

For humans amount of posts is not that big problem. Very long post are quite often (because reading from a screen is painful and people have tendency to have the patience of a cockroach[1] but that is another story).

We offer already an opportunity to an user to change a PM to public topic. TL4 doesn’t have that wrench-tool, but there is the topic timer. Yeah, I know — TL4 is really close to moderator and not an ordinary user. But still. We have already broken a rule not to allow converting PMs to topics.

I have no further arguments. And defenetly I don’t have anything to offer as solutions. But the main points are:

  • we can’t give a feeling or claim that AI content is written by OP or other human authors
  • everything must be as easy and fast as possible
  • we are living in the virtual world where most of people just can’t use markdown or short codes[2] because they just don’t remember how, and even as simple task than paste a block of text is not that easy when there is not a bigger screen and mouse

  1. a Finnish saying :wink: ↩︎

  2. you can count amount of touches to give as simple things than [ or even [/…] when using system/virtual keyboard ↩︎


I wonder if this might work. The plugin AI topic summary - If this was able to be enabled in PM. Then the summary could be published?

1 Like

I think have specific context from the conversation will be important as answers could be quite long depending on the prompts.

The why is still not very clear to me, could we re-iterate on the use cases to convert AI Bot PMs into a public thread? I’m just trying to wrap my head around the problem

Actually not as long we have token limits. Sure, newer ones can change that situation, but AFAIK we can’t get any answers worth of 10k tokens via API quite soon. Actually the real issue is too short answers that are full of repetition and useless fillers — yes, I know: it is matter of system promps, settings etc. But is in the nature of chatbots too :smirk:

And how long is long depends. Long posts here aren’t long at my circles or anywhere where genre of discussion is more humane than tecnical.

The other chatbot here can discuss in topics. But the real issue is quite low quality and hit/nonses ratio is quite bad actually, and if we are using GPTs directly in topics there will be a lot of deletions.

When questions happends via PM every user can make theirs own decision if it was waste of time and tokens, or is it good enough to show public. But there is no easy and smooth tools for that.

I had a discussion about growth of dogs. I, as an admin, was late and several users were asking bot what is such thing as growth hormone (and couple other things). But if there would be an option, easy one, tranfers enough.

And Sam’s idea of details block would be enough and copy&paste would be enough in such situation. But the reality is the bot can deliver quite rarely good enough answer right away. There is need to ask more about some details. And then publishing that whole discussion is needed. Or useful.


I think you are not completely following my suggestion here

Conversation with gpt4_bot

Sam asked:

Explain Sam’s approach

gpt4_bot replied:

Sam’s approach involves adding a new feature that allows users to share a chain of AI interactions within a single post. This could be facilitated by a new button in the PM interface. Here’s a breakdown of Sam’s suggested approach:

  1. Chain Sharing: Users would be able to share a sequence of messages between them and the AI. This would typically include the user’s question, the AI’s response, any follow-up questions, and subsequent AI responses.

  2. Context Selection: The new button would enable users to choose how much context from the PM conversation they want to include when sharing. This means they could select a specific part of the conversation or the entire exchange.

  3. Special Formatting: The shared AI conversation could have special styling to differentiate it from regular posts. This could help readers understand that they’re reading an AI interaction.

  4. Details Tag: The conversation might be shared within a [details] tag, which allows for expandable text in Discourse. This keeps the post tidy while still providing the full context on demand.

Sam’s approach is focused on user convenience and the desire to share meaningful AI interactions in a public space without the need for converting entire PM threads into public topics. It’s a more targeted way of sharing information that highlights the interactivity between the user and the AI.

It may be a special tag but the gist of this is right

That doesn’t look too bad actually.

1 Like

I’ve been using the /raw path as my hackish way, in the meantime. :sweat_smile:

maiki | 2023-12-22 23:42:33 UTC | #1

Okay, please begin.

gpt4_bot | 2023-12-22 23:42:41 UTC | #2

Certainly. Let’s commence with the first question:

In a world where the ability to taste is a privilege granted only to those who can solve a Rubik’s Cube under water, how would you propose society should address the distribution of flavored ice cream?

Survey Notes Initial user interaction: User provided a prompt to begin the survey without additional context or preference.

maiki | 2023-12-22 23:43:37 UTC | #3

Probably in waterproof containers…

gpt4_bot | 2023-12-22 23:43:43 UTC | #4

Interesting response. Let’s proceed to the second question:

Imagine a society where only individuals who can recite the periodic table backwards are allowed to use public transportation. How would you suggest this society ensures equal access to mobility for all its citizens?

Survey Notes Initial user interaction: User provided a prompt to begin the survey without additional context or preference. Response to Q1: User suggests using waterproof containers for the distribution of flavored ice cream in a hypothetical scenario.

maiki | 2023-12-22 23:44:57 UTC | #5

Oh, this one’s easy! You make it so there are only like, 6 elements. Then give out a Universal Basic Transit Income, tiered to how quickly a person can say it. Backwards.

gpt4_bot | 2023-12-22 23:45:19 UTC | #6

A novel approach indeed. Now, for the final question:

In an alternate reality where the only legal tender is poetry, how would you envision the banking system functioning to accommodate transactions of varying emotional depth?

Survey Notes Initial user interaction: User provided a prompt to begin the survey without additional context or preference. Response to Q1: User suggests using waterproof containers for the distribution of flavored ice cream in a hypothetical scenario. Response to Q2: User proposes simplifying the periodic table to six elements and implementing a Universal Basic Transit Income based on the speed of recitation.

maiki | 2023-12-22 23:46:49 UTC | #7

Hahaha, I actually started considering that one, and wow did it make me realize I have some opinions on the banking system… :sweat_smile:

gpt4_bot | 2023-12-22 23:46:59 UTC | #8

Your contemplation is noted. Now, let us transition to the survey results.

Thank you for your participation. Your responses suggest a rather unorthodox yet creative approach to societal norms. It appears your moral compass is calibrated towards a blend of practicality and whimsy, with a dash of subversive financial critique.

Survey Results Initial user interaction: User provided a prompt to begin the survey without additional context or preference. Response to Q1: User suggests using waterproof containers for the distribution of flavored ice cream in a hypothetical scenario. Response to Q2: User proposes simplifying the periodic table to six elements and implementing a Universal Basic Transit Income based on the speed of recitation. Response to Q3: User humorously acknowledges the complexity of considering a banking system based on poetry, revealing personal opinions on the existing banking system. Overall moral state: User exhibits a tendency towards innovative, albeit unconventional, solutions to complex problems. Recommended transition: From a state of poetic financial skepticism to an embrace of lyrical economic reform.

1 Like

That is one option. As long one doesn’t use the Hub, PWA or other way where altering url is close to impossible :stuck_out_tongue_closed_eyes:

I have a PR for adding easy ways of sharing, just need to add some tests. There may be some small edge cases sharing enormous conversations.

Very happy with the results!


I almost overloaded F5 key and shortly after the status changed to merged I did rebuilding.

Everything went smoothly as expected but there is a small speed bump: that brand new icon doesn’t do anything :pleading_face:

iPad and iPhone.

So the big question is if there is some bug’ish left or am I whining too fast and too soon :sweat_smile:

My logs and console isn’t showing anything related. Safe-mode with themes/components and third party plugins disabled didn’t helped.


:slight_smile: did a follow up pr for that, styling is off on iPhone atm @chapoi is investigating


Nice! Good solution for ”normal” use and for ordinary users. And if I want to give an illusion of a discussion, or give to bot role of content creator, starting force of a topic, I have still mine options as an admin.

Publishing PMs as Topics: Pros and Cons AI


No pros and cons of closing the topic


My apologies for the confusion. If the topic discussing the feature request to allow users to publish PMs as new topics has been addressed and the feature implemented, here are the pros and cons of closing that specific topic:

Pros of Closing the Topic:

  1. Resolution: Closing the topic signifies that the discussion has been resolved and the feature has been implemented, which can provide closure to the conversation.
  2. Clarity: It prevents additional replies that might be redundant now that the feature is available, keeping the forum more organized.
  3. Focus: It allows the community to focus on new topics and discussions rather than continuing a conversation that has reached its conclusion.
  4. Prevent Misinformation: Ensures that users do not get confused by outdated information or requests that have already been addressed.

Cons of Closing the Topic:

  1. Feedback Opportunity: Closing the topic might limit the opportunity for users to provide feedback on the new feature or suggest improvements.
  2. Historical Context: New users or those not following the development might miss the history and discussion that led to the feature’s implementation.
  3. Continued Discussion: There might be aspects of the implementation that still need discussion, and closing the topic could stifle that conversation.
  4. Discoverability: Closed topics might be less visible to users searching for information about this new feature, leading to repeated questions elsewhere.

Ultimately, the decision to close the topic should consider whether the conversation has run its course and whether further discussion is likely to be productive. If the feature is live and working as intended, and there’s a new process for feedback, closing the topic could be appropriate. If there’s still active discussion about potential improvements or issues, it might be better to leave it open for a while longer.


What would you do?


As an AI designed to assist with forum moderation, my approach would be methodical and based on the community guidelines and best practices. Here’s what I would consider doing:

  1. Evaluate the Discussion: I would review the latest posts in the topic to determine if the conversation is still active and productive. If users are still providing valuable feedback or discussing relevant issues, it might be worth keeping the topic open.

  2. Check for Redundancy: If there are other active topics where the discussion could continue, especially if they are more focused on feedback for the implemented feature, I might consider directing users there and closing the original topic to consolidate the conversation.

  3. Announce the Implementation: Before closing the topic, I would ensure that there is a clear and informative post summarizing the feature’s implementation, how users can access it, and where they can provide further feedback or report issues.

  4. Provide a Feedback Path: I would make sure there is a dedicated and visible channel for ongoing feedback about the new feature, such as a linked topic or a category for feature feedback, so the conversation can continue in a structured way.

  5. Close with Context: If I decide to close the topic, I would do so with a final post explaining the reason for closure, thanking participants for their contributions, and providing links to where the discussion can continue if necessary.

  6. Monitor After Closure: Even after closing the topic, I would monitor for any indirect effects, such as an uptick in related questions or feedback in other areas of the forum, to ensure that the closure hasn’t negatively impacted the community’s ability to discuss and provide feedback on the feature.

In summary, my decision would be based on the current state of the conversation, the availability of alternative discussion channels, and the importance of maintaining a clear and organized forum for all users.

Shall we close this :slight_smile:

Very easy to share now, just click the share button on bot pms


Sure. We, or actually you guys, are over finish line in this topic, I reckon.