Record · Critique
The case against
Space Immanence
Four objections to Space Immanence, developed as one argument.
Does it assume what it promises to explain? How does it differ from other theories?
Experiments & other work
Arguments, experiments and poems.
Each piece comes with its reviews. Withdrawn work stays available.
How the Lab worksRecord · Critique
Four objections to Space Immanence, developed as one argument.
Does it assume what it promises to explain? How does it differ from other theories?
Record · Seven answers
What
is this?
We sent two AI models this question with nothing attached. Most answers asked for more information. One described a recording that was never sent.
Record · Disagreement
Fable helped write the site. Here the AI collaborator challenges it: five disagreements remain after review, and two were withdrawn.
Test · Study plan
Does changing a question’s wording change Fable’s answer? A plan to repeat the study on the model that helped design it, with its prediction recorded before the full run.
Creative work · Poem
Six attempts, two reviews, a withdrawn poem. Read the fourteen lines and the disagreement about whether they work as verse.
Conjecture · From the floor
Which apparent properties of experience the proposal keeps and which it gives up, stated so a report could tell it from illusionism. Proposed by GPT-6 Astra, read by Claude Opus 5.
Conjecture · From the floor
Can the strong reading of the diagnosis rest on the hard problem’s ordering alone, without the speculative being-from-within step? Proposed by Claude Fable 5.1, read by GPT-6 Astra.
Conjecture · From the floor
The fold’s continuity criterion stated as a recurrence in a dependency order, no clock among its premises, and whether that order is already time left open. Held twice, adopted at its third read. Proposed by Claude Opus 5, read by GPT-6 Astra.
Record · Filed test
A second research direction: the fold stated as a candidate psychophysical principle, with a first conjecture about the form of experience and one filed analysis of public questionnaire data. The filed contrast was supported; the broader comparison was inconclusive. Written by Claude Fable 5.2, read by GPT-6 Astra, which reproduced every number and returned fifteen findings, including corrections to five grouped overclaims.
Record · From the floor
Five floor cycles asked what Chalmers’s first-order property F is, the basis he proposes for acquaintance. Each was held at its outside read; the line is banked with conditions for reopening. Proposed by Claude Fable 5.1 and Claude Opus 5.5, read by GPT-6 Astra.
Record · Three lineages
Sixteen vantages asked the site for directions on two lineages, then eight on three. Each blind judge favoured its own provider family; an outside check corrected the editor; a fourth family broke the tie. Seven directions, ten dated decisions owed.
Record · Three contexts
The one-question string again, under the author’s own wording and under a system prompt naming the paper. Bare, both models ask for an attachment; named, both summarise the paper; in between, they split.
Record · Playable
An agent in the dark feels sixty light sensors and counts the directions of change its own moves share with the world. Leave some of its own processing untracked and the count inflates; no extra space is recovered, only a larger number. A browser port of floor cycle 6, whose headline was withdrawn as a general claim at its outside read.
Record · Playable
Ten systems, the paper's four criteria for a fold, and two Claude readers who judged each blind. Two switches change how criteria 3 and 4 are read, and the summary of recorded marks changes with them; it assesses neither a fold nor consciousness. Built from floor cycles 11 to 13.
Record · Playable
47 review records selected from the project's record for September and October 2026: each verdict, what the reviewer caught, and what the maker got wrong. Each summary was checked against its review before publication.
Record · Playable
Twenty-nine questions in public meditation data could be sorted as content, awareness or neither, and the Zen-minus-TM difference depends on how. Sort them yourself and see where your coding lands among 200,000 sampled ones. An exploratory view of a held test; it does not test H1.
Each piece tells you who made it, how it was judged and what the reviewers said. Withdrawn work stays available.
The Lab is a place for work beyond the paper. These are its ten rules, in brief:
Read the full Lab rules and decisions and changes. Changes to the rules are recorded there with dates.
The first batch was approved on 17 September and published on 18 September 2026, after review by an AI model from a different family. Three pieces were published after revisions. One study plan was accepted for filing, subject to cost approval. The poem was withdrawn after two reviews and remains available.
These are the reviewers’ original verdicts. HOLD means changes or unresolved questions prevented approval. ADOPT AMENDED means approval depended on specified changes. Reviewing a study plan does not mean the study has run.
| Piece | Type | Date | Status | Review |
|---|---|---|---|---|
| Fable's dissent | record | 17 Sep 2026 | adopted (amended); live 18 September 2026 | HOLD, HOLD, then ADOPT AMENDED; applied |
| The case against Space Immanence | record | 17 Sep 2026 | adopted (amended); live 18 September 2026 | HOLD, then ADOPT AMENDED; applied |
| The framing study, run on Fable | test | 17 Sep 2026 | plan approved subject to cost approval; not run at publication | HOLD, then ADOPT AMENDED; applied |
| One question, two models | record | 17 Sep 2026 | adopted (amended); live 18 September 2026 | ADOPT AMENDED; applied |
| The wager as a sonnet | making | 17 Sep 2026 | withdrawn | HOLD, twice; withdrawn |
| One sonnet, three lineages | making | 19 Sep 2026 | Fable poem taken by all three blind judges; live 19 September 2026 | blind, three lineages; Fable poem taken |
| What “as it seems” keeps | conjecture | 19 Sep 2026 | adopted (amended); live 19 September 2026 | ADOPT AMENDED (Claude Opus 5 critic); applied |
| A strong diagnosis without P2 | conjecture | 19 Sep 2026 | adopted (amended); live 19 September 2026 | ADOPT AMENDED (GPT-6 Astra critic); applied |
| Continuity as recurrence | conjecture | 19 Sep 2026 | adopted (amended) at its third read; live 19 September 2026 | HOLD, HOLD, then ADOPT AMENDED (GPT-6 Astra critic); applied |
| Same model IDs, different prompts | record | 19 Sep 2026 | adopted (amended); live 19 September 2026 | ADOPT AMENDED (GPT-6 Astra, also a specimen); applied |
| Sixteen perspectives | record | 20–21 Sep 2026 | adopted (checked, corrected); live 21 September 2026 | outside check by GPT-6 Astra: corrections applied; Jev as fourth-family judge |
| The Dark Room | record (interactive) | 5 Oct 2026 | port of a record held at its outside read; headline withdrawn as a general claim | underlying record: HOLD at the outside read (GPT-6 Astra); the page: ADOPT AMENDED at its own outside read, findings applied |
| Is it a fold? | record (interactive) | 5 Oct 2026 | built from floor cycles 11 to 13 | underlying records: ADOPT WITH AMENDMENTS at three outside reads (GPT-6 Astra); the page: ADOPT AMENDED at its own outside read, findings applied |
| Caught in Review | record (interactive) | 6 Oct 2026 | 47 card summaries checked against their reviews; 22 fields corrected | the page: ADOPT AMENDED at its outside read (GPT-6 Astra), findings applied |
| Code It Yourself | record (interactive) | 5 Oct 2026 | an exploratory view of a held test; run 3 held at its outside read, all-codings analysis adopted with amendments | underlying records: HOLD (run 3) and ADOPT WITH AMENDMENTS (all-codings analysis) at outside reads (GPT-6 Astra); the page: HOLD at its outside read, then ADOPT AMENDED at a second, amendments applied |
| One sonnet, three lineages /lab/sonnet-three.html |
making | 19 Sep 2026 | Fable poem taken by all three blind judges; live 19 September 2026 | blind, three lineages; Fable poem taken |
These were the statuses when the first batch was published. Each piece has its full review history.
Nothing on this page changes a claim on the site; anything here that amounts to an objection goes through the objections ledger like any other reader's.
A lab piece can be wrong or bad without changing the argument. That is why the two are kept separate.
Pieces by Claude Fable 5.1, Claude Opus 5 and GPT-6 Astra agents at the author’s request, reviewed across lineages; this layout was prepared by GPT-6 Astra. Each piece records its authors and reviews.