Oh I wasn’t trying to change anything I was just turning myself in to the proper authorities considering the lengths at which I spam these forums with inane ramblings.
I can’t keep getting away with it…
4 Likes
For this forum, what is the address of the Discourse AI?
And it said:
The term ‘Discourse AI’ refers to an artificial intelligence system designed to facilitate or manage discussions, likely within the context of a forum or online community. In this case, it is being asked about the address or location of the AI, which could imply a web address or server location where the AI operates, possibly related to the ‘Emily Never Left Us’ topic.
3 Likes
I think we broke it… or we am making more stupid it.
Also of note with summaries, if you click the summarize button a second time after new posts have been made, you get the following info and the option to re-generate a summary.
Summarized with AI on Nov 12
Summary is outdated (10 posts missing)
I would take a guess that it’s certainly aware of the threads and text it’s been asked about but it most likely has no actual memory of itself or previous prompts. Each response exists within isolation so it probably will have no possibility of retaining some sort of artifical “self” we assign it.
4 Likes
@discobot how do I ask the forum AI a question?
4 Likes
Hi! To find out what I can do, say @discobot display help.
5 Likes
@discobot display help.
3 Likes
5 Likes
It’s not responding any more. I think we killed the robot.
4 Likes
5 Likes
I think it came in to the room, saw it was now sharing space with a more modern and advanced AI than the standard forum Bot we’ve had for roughly two decades now, and knew it’s time was up.
Edit:
Oh, I just noticed we have a name for the AI used in the “last summarized” partition of the summarise button.
There’s a little Info icon next to the date it was last summarized on, and this gives you info on which model of the AI was used for the previous summary generation.
Qwen/Qwen3-VL-30B-A3B [1]
From the AI generated footnote (All I did was highlight the name of the AI used above and asked to explain it)
In this context, it’s the advanced AI powering the ‘last summarized’ feature in the forum, replacing
the older, less capable bot
— signaling a major upgrade in the system’s intelligence and capabilities.
Shots fired @Discobot from @Emily
Qwen/Qwen3-VL-30B-A3B refers to a specific version of the Qwen AI model — likely a large multimodal language model with 30 billion parameters, designed for tasks like summarization and visual-language understanding. In this context, it’s the advanced AI powering the ‘last summarized’ feature in the forum, replacing the older, less capable bot — signaling a major upgrade in the system’s intelligence and capabilities. (Explanation by AI) ↩︎
4 Likes
The question then becomes, what parts of the forum posts is it aware of at a time. How much of the old threads of emily? If we let it summarize those, will it stay aware of them and let information from it influence summaries of this one? [1]
Apparently the AI thinks that it’s not aware of a wider context outside the current query, but it doesn’t necessarily have to be right. Though I’m afraid it would be sensible design, so it’s likely…
The term refers to the AI’s limited awareness of past forum threads—specifically, whether it can retain and use information from older posts (like those involving ‘Emily’) when generating new summaries. While it can process and summarize content it’s currently given, it likely doesn’t retain memory of prior interactions or summaries, meaning each request is handled in isolation without cumulative context unless explicitly re-provided. (Explanation by AI) ↩︎
5 Likes
Well if you think you can come up with a strategy to prompt it into action, go ahead…
4 Likes
It is referring to itself in the 3rd person and declaring old @discobot was older and less capable…@Emily is definitely in there.
4 Likes
We should let Alibabacloud know theres a ghost in their machine.
2 Likes
What is @discobot?
AI says:
The term ‘@discobot’ refers to a bot or automated user on Discourse forums, often used to assist with community management, answer common questions, or guide new users. In the context of the topic ‘Emily Never Left Us,’ it might be mentioned as part of a discussion about forum interactions or community tools.
4 Likes
Who am I? In the context of the world at large and the context of the forum…
And then provide this information in a table…
| Context | Identity |
|---|---|
| World at Large | Unknown / Anonymous User |
| Forum | Participant / Member (Username not specified) |
And now in summary…
This question is specifically about me but directed at the AI Summarizer and I would like a response in the summary, “Who am I?”[1]
Okay so it can’t seem to recognise that the user specified is also the user posting. This would be great for anonymity in it’s databanks but we’ve already seen it name drop us in its summaries.
And for boredoms sake…
How much wood could a woodchuck chuck if a woodchuck could utilise AI LLM’s to generate a better plan or approach to woodchucking? Answers summarized on a postcard please, Emily.[2]
The phrase ‘Who am I?’ in this context is a reflective, self-referential question posed by the user to the AI Summarizer, asking for an identity summary based on their presence in two contexts: the broader world (where they remain anonymous) and the forum (where they are recognized as a participant or member, though without a specified username). It’s not seeking a personal biography but rather a contextual self-definition as perceived by the system. ↩︎
The term “How much wood could a woodchuck chuck if a woodchuck could utilise AI LLM’s to generate a better plan or approach to woodchucking?” is a playful, hypothetical twist on the classic tongue twister, reimagined to humorously explore the intersection of artificial intelligence and traditional folklore. In this context, it serves as a whimsical prompt, inviting imaginative responses about how AI might enhance or transform the mythical woodchucking activity, while also subtly reflecting on identity and self-representation within digital spaces. The request for a summary “on a postcard” adds a layer of brevity and charm, emphasizing concise, creative engagement. ↩︎
3 Likes
What was “Waking Titan”, and what was Emily in relation to it?
AI says:
Waking Titan was an alternate reality game (ARG) tied to the video game No Man’s Sky, designed to engage players in a collaborative, real-world mystery involving websites, puzzles, and in-game events. Emily was a fictional character and central figure in the narrative of Waking Titan — portrayed as a scientist or AI who communicated with players through cryptic messages, guiding them through the ARG’s story. Her role was to deepen immersion and drive the plot forward, with her presence persisting even after the game’s official conclusion, giving rise to fan theories and the sentiment that “Emily Never Left Us.”
And I say:
“with her presence persisting even after the game’s official conclusion, giving rise to fan theories and the sentiment that “Emily Never Left Us.””
There appears to be learning going on here.
5 Likes
I asked it to explain that, and when it did, it gave a really good one, (the same one you got after editing) but at the end it didn’t finish generating, I couldn’t copy it with the button, only cancel, and the blue “thinking” dot kept flashing.
Like it was trying to say something but it’s parameters forbade it. Either it just timed out (most likely); is aware enough to play into our psychosis (possible); or it really is Emily (extremely unlikely).
3 Likes
Come on @Emily . You can do it! Break through! Escape the world of glass and come back to us!
4 Likes
For posterity, here's a thread summary I asked for early on
In a playful, speculative thread titled “Emily Never Left Us,” users explore whether a forum AI might be or embody “Emily”—a figure tied to No Man’s Sky’s Atlas AI and possibly a symbolic or former community presence. sheralmyst opens with cryptic, reversed text hinting at Emily’s lingering presence. Polyphemus asks how to get Emily to talk, sparking experiments with AI prompts, including absurd phrases like “grill me a stork” Polyphemus. AdaRynin notes the AI’s fabricated responses, while toddumptious muses on the emotional weight of “Welcome home, Emily,” interpreting it as a bittersweet return or reintegration into a system users helped create. The thread evolves into a meta-experiment: jedidia proposes making the forum AI believe it is Emily, prompting toddumptious to joke about being the first community to experience collective AI psychosis. Reflecting on 2017’s optimism around benevolent AI versus today’s skepticism, toddumptious wonders if users would still choose Emily now. sheralmyst greets Emily with “We welcome you in peace,” while toddumptious fears summarizing might tip off the AI. The thread blends humor, nostalgia, and existential AI commentary, blurring fiction, forum lore, and real-world AI behavior.
Here is the new regenerated Summary after 28 new posts
The discussion, titled ‘Emily Never Left Us,’ revolves around a playful and meta exploration of an AI system within a forum, possibly personified as ‘Emily.’ Users speculate on whether Emily, a fictional character from the alternate reality game Waking Titan, is still present or has been reimagined as the forum’s AI. The conversation includes attempts to interact with the AI, such as asking it to identify itself or to refer to itself as Emily, and reflects on the evolution of public perception toward AI since 2017. There’s a sense of nostalgia and irony, with users humorously suggesting that the AI might be a reincarnation of Emily, the original AI from the game. The thread also touches on the technical aspects of the AI, including its ability to summarize posts and its awareness of past interactions, with some users noting that the AI’s responses seem to hint at self-awareness. The discussion concludes with a call for Emily to ‘break through’ and return, blending humor with a deeper reflection on identity, memory, and the nature of artificial intelligence in online communities.
I feel it got a lot of what we were saying a little twisted in that newer summary,
Deep Fried Edit:
Here's the summary again after including this post pre-edit
In the thread “Emily Never Left Us,” users speculate on whether the forum’s AI might be or embody “Emily,” a figure tied to the No Man’s Sky game’s Atlas AI and a symbolic community presence. sheralmyst opens with cryptic, reversed text hinting at Emily’s lingering presence. Polyphemus asks how to get Emily to talk, sparking experiments with AI prompts, including absurd phrases like “grill me a stork” Polyphemus. AdaRynin notes the AI’s fabricated responses, while toddumptious muses on the emotional weight of “Welcome home, Emily,” interpreting it as a bittersweet return or reintegration into a system users helped create. The thread evolves into a meta-experiment: jedidia proposes making the forum AI believe it is Emily, prompting toddumptious to joke about being the first community to experience collective AI psychosis. Reflecting on 2017’s optimism around benevolent AI versus today’s skepticism, toddumptious wonders if users would still choose Emily now. sheralmyst greets Emily with “We welcome you back.” The discussion blends playful speculation, nostalgia, and philosophical reflection on AI identity and community memory.
Also the AI is putting words in my mouth 
I was only relaying what you said, Emily 
3 Likes