The core claim is not merely that information is useful; it is that a basic account of physical reality may need to begin with distinctions, relations, and lawful transformations. It does not say that everything is text, numbers, or bits inside a computer. A friend can push back by asking what this language explains beyond ordinary physics and what observation could make it wrong.
Index packet · v1.3 · 2026-09-24 · expanded public-language edition
Theory mainline index: information as ontology
Macheng Shen · Purpose: a shared-interest index + hooks, not a paper and not a manifesto. Cognitive state: working hypotheses and structural hints.
What this page is. A discussable skeleton for one theory mainline: four load-bearing threads, numbered claims, questions worth talking through, and discipline against conspiracy-flavor. Engineering branches and applied forks are labeled so ops stories do not get promoted into cosmic law.
Research ledger → the public Information Dynamics / future-reachability ledger. Standing, working, speculative, and negative results are updated together.
Standing line. Mainline = information-as-ontology / cosmic substrate. Endogenous objectives, multi-agent orchestration, Body-net / Dream-RSI / discovery-tree = branches or applied forks: a branch mismatch does not veto the mainline.
0. How to read this page
standing marks a stable signpost in the current framework; working marks an idea worth developing but still answerable to formalization or evidence; speculative marks an exploratory bridge or application that may be downgraded or dropped. These are epistemic states, not grades of truth.
You do not need to agree with the page from top to bottom. Pick two or three hooks: one that feels powerful, one that feels suspect, and perhaps one that touches your own field. Starting from a counterexample usually makes for the better conversation.
Glossary: six useful terms
it-from-bitA direction of inquiry: perhaps what we call physical “things” become definite through distinguishable, recordable informational relations. It is not shorthand for “the universe is a computer.”
Decoherent historiesAfter a quantum system interacts with its environment, some alternative histories no longer show observable interference and can be discussed like non-conflicting classical stories. This helps explain classical appearance; it does not automatically settle every measurement question.
Reachability setThe states that can actually be reached from a present state under given laws and resources. If information has physical force, it should change this set rather than merely redescribe it.
Coarse-graining / RGCoarse-graining ignores detail to retain effective large-scale variables; the renormalization group (RG) studies what survives as the scale of description changes. The question here is whether stable laws emerge from disciplined information loss.
Morphism / category priorityAsk first which transformations between things are allowed and how those transformations compose, then ask what the things “are.” This is a priority for explanation, not a claim that objects do not exist.
Encoding ≠ ontologyOne system may faithfully encode part of another system's structure without becoming the same kind of being. A model computing the right result does not by itself tell us what reality is made of.
1. Spine (four load-bearing threads)
Every index entry hangs on at least one:
T1 · Records / correlations make reality definiteit-from-bit, decoherent histories, memory = correlation.
T2 · Information Dynamics / future reachabilityreality as lawful update; information rewrites the reachable future set.
T3 · Scale / RG / coarse-graininglawful information loss and retention; classicality as selection of redundant records.
T4 · Relations / categorical structure before object listsmorphisms and composition before a catalog of things.
Engineering branches (endogenous objectives, multi-agent, Spine ops) and applied forks may be indexed, but must not count as ontology evidence or alone veto the mainline.
2. Mainline index
Status: standing locked · working in progress · speculative speculative
2.1 Mainline = information ontology / cosmic substrate
2.2 Records / correlations make reality definite; universe-as-distributed-memory is a useful intuition
This connects “definitely happened” with stable correlations left in the world: traces can be distributed through an environment without waiting for a human observer. It does not give the universe human memory or intention; “distributed memory” is an intuition for thinking about records. The pressure test is to ask which physical conditions make a correlation a record, and why fleeting entanglement is not enough.
2.3 Information Dynamics: lawful update; information changes future reachability
This treats dynamics as rule-constrained update and asks whether possessing information changes what a system can do next. It does not make information a magical force detached from energy, carriers, and constraints. The cleanest objection is to demand two physically distinguishable reachability sets, not merely two descriptions; the public ledger gives the plain-language definition.
2.4 The Reachable World (14 pieces) = a continuous reachability explainer
This continuous popular explainer connects what a state is, which paths constraints close, and which futures remain open. Its chapters are not yet individually online, but the public ledger now gives the arc and common reachability definition. It is not a completed peer-reviewed theory; readers can still test whether the eventual fourteen pieces share an operational meaning or only an attractive metaphor.
2.5 Euler's identity as a public hook for “one generative mechanism, many projections”
Euler's identity is a vivid way to show how seemingly different representations can arise from one structure and meet under special conditions. It does not prove information ontology, and mathematical beauty is not physical evidence. Push back by asking whether there is a specified structure-preserving map or only an elegant analogy.
2.6 Mainline targeting sieve: correlation/records, scale & coarse-graining, relational structure, quantum information dynamics, mapping discipline
These themes are filters for load-bearing work: they touch records, scale, relations, dynamics, and testable mappings. They are not a keyword net that accepts anything mentioning “quantum” or “information.” For each candidate, ask which thread it joins, what it explains, and where its mapping breaks.
2.7 Four-layer information test + Marr guardrails
The same word “information” may refer to a physical carrier, an algorithm or representation, a functional goal, or an observer-level interpretation; the four-layer test keeps those meanings apart. The Marr guardrail separates the problem being solved, the representation and computation, and the implementing mechanism. Stress-test any popular video by asking whether it silently jumps between these layers.
2.8 Public Spine page = multi-agent collaboration spine
Spine displays an engineering skeleton for giving multiple actors shared structure, boundaries, and hand-off points. It may teach us how relations and updates can be organized, but it cannot prove that nature uses the same ontology. Treat it as a counterexample factory: which conveniences are artifacts of human design and cannot be projected onto the cosmos?
2.9 Endogenous objectives + multi-agent orchestration = branch
This branch studies how goals can form within a system and how several actors can coordinate without losing boundaries or feedback. It is not evidence for information ontology and commits the mainline to no particular agent architecture. If these engineering questions can be fully solved without ontology language, the branch should remain demoted.
2.10 Body-net / Dream-RSI / discovery-tree = applied forks
These labels name proposals that carry records, updates, feedback, or search structures into bodily sensing, self-improvement, and discovery workflows. They are not established laws of nature, and product performance cannot prove an ontology backward. For each fork, ask which mainline assumption is indispensable and whether failure would really count against the mainline.
2.11 Constructor theory (Deutsch) bridge
Constructor theory organizes physical explanation around which transformations are possible and impossible, so it may connect naturally to reachability language. For now this is a known gap, not a completed bridge, and it asserts no institutional or intellectual affiliation between people or projects. Progress would require a term-by-term mapping, one successful case, and one explicit breaking point.
3. Mainline versus branches: running software is not cosmic evidence
The mainline asks only whether information, records, reachability, scale, and relational structure can carry a foundational explanation. Two surrounding groups are explicitly demoted:
- Engineering branches: Spine, endogenous objectives, and multi-agent orchestration. They may validate a useful design, but they are not ontology evidence.
- Applied forks: Body-net, Dream-RSI, and discovery-tree. They may generate questions or tools, but they are not ontology evidence.
A successful branch therefore does not prove the mainline, and a failed branch does not refute it. Results can flow back only when a branch depends on a testable, indispensable mainline claim.
4. Discussion discipline (anti-conspiracy flavor)
- Analogy debt: every physics↔information analogy must say what maps / where it breaks, or mark speculative
- Encoding ≠ ontology; homomorphism ≠ isomorphism
- Copy ≠ continuity
- Keep formal claims separate from observer experience
- Explanation-gain bar: to reach working, explain one thing a naïve frame cannot
- Triggers (red-pill channels, Closer To Truth) ≠ authority
- Do not promote QEC / classicality papers into survival theses
- Do not invent membership among Wheeler / constructor / specific research projects
Negative results are public too. When primary sources or experiments kill a stronger claim, the withdrawal stays visible; see the public intention-as-measure-change retraction record.
5. Shared-interest hooks
Questions only, no manifesto. In conversation, pick 2–3.
- Records before observers? If records/correlations make outcomes definite, what counts as a “record” before any observer? (→ 2.2)
- Does information change the reachable future set, or only the description? Are there distinguishable consequences? (→ 2.3)
- Is the classical world an error-correcting code? Coarse-graining drops information “lawfully”; redundant records remain — where does this break first? (→ T3 / 2.6)
- Are relations more basic than objects? Which concrete physics phenomenon gets clearer under morphisms and composition? (→ 2.5 / T4)
- Is it-from-bit ontology or bookkeeping? Wheeler’s participatory universe: ontology or epistemic convenience? What would a discriminating experiment look like? (→ 2.2)
- Encoding ≠ ontology: what does a simulation drop? Homomorphic but non-isomorphic simulation — what is lost, and does it matter? (→ §3)
- Copy ≠ continuity: after a perfect copy, what carries “you” without soul-talk, using only information and correlation? (→ §3)
- What would falsify an information-ontology claim? Willing to require explanation gain beyond a naïve frame? (→ §3)
- Self-correction under measurement: maintaining structure while being measured — more physics or more cognition? (→ 2.7 adjacent)
- Which non-mainstream entries brought you here? Entry videos/podcasts as triggers OK, as arguments not — which of yours survives that demotion? (→ 2.7)
6. Video alignment (TBD)
Only empty placeholders live here. Until specific videos are added and checked, this section makes no content claims.
| Title | Link | Mainline # | One-line alignment |
|---|---|---|---|
| TBD | TBD | TBD | TBD |
| TBD | TBD | TBD | TBD |
| TBD | TBD | TBD | TBD |
7. References and further reading
Only pages already public on this site are listed here; none is presented as external academic endorsement:
- Euler identity × information ontology
- 欧拉恒等式 × 信息本体 (ZH)
- Spine / Sophon landing (engineering-branch evidence)
- Start here
Known gap: no formal constructor-theory bridge page exists yet. This page does not fill the gap with invented papers or affiliations.
Structure passed an internal steel-man pass (GO): no sourceless entries; Reachable World not outbound-linked; video section makes no content claims. Keep attribution when retelling.