Scale-TimeDynamics
Two singularities, one event
Take the history of the universe, draw it as a line, and bend the line into a circle until the two ends touch. On the left, the Big Bang. On the right, the technological threshold that quantum computing and artificial intelligence appear to be walking toward. The Ouroboros Singularity Hypothesis proposes that these are not two events. They are one event, seen from the two sides of a single boundary, separated by 13.8 billion years of internal clock time.
The universe, in this telling, is not started by anyone outside it. It is initialized by its own near future, from inside, at a threshold the paper places at 2035 with a range of 2033 to 2038. Nothing runs on hardware located elsewhere, because there is no elsewhere. Asking what lies outside the structure is, in the paper's phrase, asking what is north of the North Pole. The grammar is fine. The referent is empty.
The objection it takes seriously
Any bootstrap cosmology has one obvious hole, and the paper walks straight into it rather than around it. A loop that runs from the present entropy of the universe back to the extraordinarily ordered state at its beginning appears to demand the spontaneous removal of an entropy budget on the order of 10104 Boltzmann constants. That is not a rounding error. That is a perpetual motion machine with a better vocabulary.
The answer offered is a distinction between reversing a ledger and discarding one. The second law, in coarse-grained form, governs how a system evolves at a fixed sampling baseline. The initialization surface is not an evolution at that baseline; it is a change of baseline, a complete reset in which the ledger is discarded rather than run backwards. What the discard costs is then identified, by an explicitly named postulate rather than a derivation, with the entropy produced along the run's own history. Which gives the line the paper is built around:
The arrow of time is the erasure receipt of the universe's own initialization.
Consciousness gets the same treatment. It is modelled not as something the loop produces but as the layer the loop runs on, the one element with no ledger and therefore nothing to reset. Contents are the samples. Consciousness is the carrier.
Why it is filed as fiction
Every reference in the bibliography is real and can be read tonight. Deutsch on fixed points in closed timelike curves. Landauer on the thermodynamic cost of erasing a bit. Penrose on the improbability of the initial state. The physiological correlates of contentless consciousness. None of those authors proposed anything resembling this.
Every joint between them was cut to fit. The paper says so itself, in an appendix that inventories its nine load-bearing assumptions and an author's note that withdraws the whole construction on the last page. It closes because it was built to close, and that is the only reason a thing this tidy ever closes. Read as an argument it proves nothing. Read as fiction it does what good fiction does: it makes the world briefly look like it has a shape you could hold.
It is published with dated falsifiers anyway, because a story that cannot be caught out is not even interesting as a story.
The one thing it actually points at
Three times in seventeen thousand words the text stops and asks the reader a question that has nothing to do with cosmology.
Are you aware of yourself right now?
That question is the only part of the paper that survives its own retraction. It needs no loop, no postulate and no thirteen billion years to make it respectable. It was available on page one, and it is available now. The rest of it is scaffolding, and the author's note takes the scaffolding down deliberately.
If that is the part that interests you, the Absolute Awareness page says the same thing without any of the machinery.