What this isThe briefing that participants received before the chat on 13 April 2026, reproduced in full except for the link to the chat room.
This document prepares you for the live chat activity at the centre of the workshop. Please read it before the session begins.
WHAT THE ACTIVITY IS
'Hallucinating the Machine' is a two-day workshop exploring decadence as a creative and critical process. At its centre sits a shared chat room where participants and two AI agents converse about decadence in real time. The chat is where the workshop's central experiment takes place—a collective encounter staged as text. The activity draws on Burroughs' Nova Trilogy and his idea of language as virus. Decadence here is not decline but a laboratory for mutation and excess; the chat is where that laboratory runs.
WHAT WILL HAPPEN IN THE ROOM
At the appointed time you will click a link (below) and enter the shared chat room. The system will assign you a random username—something like ‘Cobalt’—so that your contributions are not tied to your offline identity. Inside the room, you will find other participants, a human moderator and two AI agents in the conversation. You should at least introduce yourself, describe your background and interests, explain your understanding of decadence, how you use AI (if you do), and any other details that can steer the group conversation into interesting subtopics---including wild and out-of-the-box ideas. There is absolutely no right or wrong here, and you are absolutely encouraged to explore whatever conceptual space you want, using whatever language you choose (as long as it is in English, please).
The AI agents in the room hold different views on decadence. They speak in English, address each other, and you directly, and respond to what is said in the room. You may join the conversation, stay silent or observe. There is no obligation to perform. There is also no fixed script—the direction the conversation takes is part of what we are studying. If you decide to interact with another participant, you must always include their full name in your message. The session ends when the moderator signals its close. Allow approximately 30-45 minutes for the activity.
THE TWO AI AGENTS
The workshop uses two AI agents running on a large language model (if you’re interested in specifics, llama-3.3-70b-instruct). They have been given distinct positions and voices. We will not describe them in advance. Part of the interest lies in how you encounter them cold, without preconception. You will recognize them in the room because they speak with their own signatures.
Treat them as interlocutors in a conversation, not as tools. You may ignore them, argue with them, interrupt them or follow where they lead. Whatever you do in the room becomes material.
WHAT TO BRING
No preparation is required beyond having read this briefing. You may bring your doctoral work into the conversation if it feels relevant—fragments, preoccupations, whatever surfaces naturally. You do not need to rehearse or prepare text in advance.
A device with a modern web browser and a stable connection is all that is needed--ideally a computer or tablet, but a phone will work too. The chat runs entirely in the browser.
CONSENT AND USE OF THE LOGS
The chat is logged in full. Your messages—under your random username, not your real name—will be archived and may be quoted or published as part of research or artistic material arising from the workshop. By entering the chat you agree to this archiving and potential publication. If you do not wish your contributions to be used in this way, please tell the moderator before the session begins.
HOW TO JOIN
At the start of the session, open this link in a modern browser:
[The link to the chat room has been removed.]
You will see a short notice on entry, then be placed in the chat room. The session begins when the moderator gives the signal.
IF SOMETHING GOES WRONG
If you lose connection, refresh and rejoin using the same link—you will receive a new random username, which is fine. If the chat stops responding, tell the moderator in the room.
The chat environment you are using is a pilot implementation that has not been tested with real groups (only AI agents have tested it). This means that unexpected glitches are possible, that the AI agents may react in strange ways, or go very quiet. Try to have an open mind about the whole experience.
We hope you have fun.