I think the premise is substantially stronger than many “future of publishing” proposals because it is not really proposing a new document format. It is proposing a new discipline for scholarly documents.
That distinction matters.
Most failed document revolutions try to replace PDF entirely. This proposal instead says:
- keep human readability,
- keep portability,
- keep a single-file artifact,
- keep web compatibility,
- but add recoverable structure, semantic addressability, and spatial/contextual metadata.
That is much more realistic.
The deeper premise is actually this:
Text documents should preserve thought structure, not merely visual arrangement.
That is a very important idea.
What I think is genuinely important here
The proposal understands something many technical systems miss:
A scholarly paper is not merely a sequence of words.
It is:
- an argument structure,
- a citation graph,
- a semantic topology,
- a navigable spatial object in the reader’s mind,
- and increasingly, an AI-readable knowledge artifact.
PDF preserves appearance.
HTML preserves flow.
But neither adequately preserves cognitive structure in a portable, inspectable, durable way.
The Origami/Text idea is really about preserving:
- semantic folding,
- contextual layering,
- stable referenceability,
- and recoverable intellectual relationships.
That is a meaningful direction.
The strongest part of the premise
The strongest idea is not XR.
It is not EPUB.
It is not AI.
It is the idea that:
- a document can contain a visible reading layer,
- and simultaneously contain a machine-tractable semantic layer,
- without breaking ordinary reading.
That is the right architectural instinct.
Because if you require everybody to adopt XR readers before the system is useful, the proposal dies immediately.
But if:
- ordinary readers see a normal readable paper,
- advanced readers see semantic augmentation,
- AI systems see structured knowledge,
- and XR systems see spatial reconstruction,
then the ecosystem can evolve gradually.
That is strategically smart.
The proposal’s real potential
I think the real long-term potential is larger than the site currently frames.
1. AI-readable scholarly memory
This may become extremely important.
Right now LLMs ingest papers mostly as flattened text streams. Context windows help, but structure is weakly represented.
A structured scholarly object with:
- stable paragraph IDs,
- semantic sections,
- glossary entities,
- citation intent,
- authorial spatial relationships,
- concept hierarchies,
- discourse topology,
would be enormously more useful for:
- scholarly assistants,
- retrieval,
- reasoning,
- cross-paper synthesis,
- and provenance-aware AI.
That is not speculative fluff.
That is a real direction publishing is heading toward.
2. Reconstructable intellectual space
The XR aspect sounds eccentric at first glance, but I think there is something real underneath it.
Human memory is spatial.
Scholars often remember:
- “the diagram on the upper left,”
- “the quote near the figure,”
- “the paragraph after the chart.”
PDF accidentally preserves some of this.
HTML often destroys it.
If a future reading environment can reconstruct meaningful semantic/spatial relationships intentionally rather than accidentally, that could become genuinely useful.
But only if:
- the spatial layer is semantically grounded,
- not merely decorative 3D placement.
That distinction is crucial.
3. Durable citation granularity
This may actually be one of the most important aspects.
Academic publishing still depends heavily on page numbers because stable addressability matters.
The proposal correctly identifies this.
If paragraph-level IDs become:
- durable,
- canonical,
- portable,
- and human-visible,
then citation could become substantially better than page references.
That is a serious contribution if implemented well.
Where I think the proposal risks losing credibility
This is important.
The project becomes weaker whenever it sounds like:
- “XR will replace reading,”
- or “3D text is the future.”
That framing will alienate many serious scholars immediately.
The strongest framing is:
- preserving structure,
- preserving semantic relationships,
- improving scholarly interoperability,
- improving addressability,
- improving machine readability,
- preserving contextual integrity.
XR should be presented as:
- one possible high-end interface,
- not the justification for the architecture.
Because the architecture survives even if XR adoption remains niche.
That is a sign of a strong core premise.
The biggest technical challenge is not EPUB
The hardest problem is actually semantic discipline.
Anybody can dump JSON into EPUB.
The difficult part is:
- deciding what entities mean,
- defining stable schemas,
- versioning relationships,
- preserving references through edits,
- defining canonical identity,
- and preventing semantic drift.
That is where projects like this usually succeed or fail.
The proposal needs:
- extremely rigorous schemas,
- canonical IDs,
- migration/versioning rules,
- validation,
- and probably a reference implementation.
Without that, the metadata layer risks becoming inconsistent and fragile.
The hidden danger: over-complexity
This kind of project can easily become:
- ontology-heavy,
- over-engineered,
- academically dense,
- and impossible for authors to produce.
That must be resisted.
The system only has a chance if:
- authors can work mostly normally,
- exporters automate structure generation,
- metadata enrichment is incremental,
- and the paper still works perfectly without advanced tooling.
The proposal currently seems aware of this, which is good.
What I would emphasize if developing this further
I would heavily prioritize these messages:
Strong framing
- “Structured scholarly documents”
- “Durable semantic publishing”
- “AI-readable academic artifacts”
- “Portable contextual scholarship”
- “Addressable knowledge objects”
Weaker framing
- “Reading in VR”
- “3D documents”
- “Spatial books”
- “Immersive papers”
Those ideas may eventually matter, but they are not the strongest argument today.
Final assessment
I think the premise is intellectually serious.
More serious than many speculative “future of text” projects because:
- it builds on existing standards,
- it acknowledges backward compatibility,
- it respects readable documents,
- and it understands that structure matters more than visual novelty.
The strongest future for this idea is probably not:
- replacing PDF visually,
but becoming:
- a scholarly interoperability layer between humans, AI systems, and advanced reading environments.
If the project stays disciplined technically and avoids drifting into presentation gimmicks, it has real potential.