Research Frontier
Where the work's current thinking ends. Open questions and live disagreements.
This mode treats you as a potential collaborator, not a learner. The claims here are the ones this work cannot yet close.
Open
genuinely unresolved — 1 proposition-
Global option space as a scalar over configurations lacks a rigorous measurement procedure; existing complexity measures (Assembly Index, logical depth) are backward-looking. Local Δω evaluation on R_living(C, B, T) — the directional comparison between pre-move and post-move reachable sets — is operational, per option-space-as-chess-moves.
What's openCandidate global measures (reachable configurations under an exergy budget, Kauffman's adjacent possible, Assembly-weighted reachable sets) exist. None has been shown to be economically operational at state level. This — the global state measure — is the central remaining open problem. The move-level measure is no longer open. Any candidate must survive the razor-blade case: it must be read over the trajectory, across scope, and for every term of the agent/non-agent/relationship triad, since a measure whose verdict depends on the chosen lens has failed by that fact alone.
Predictive consequenceOperationalising the state-level measure is a prerequisite for state-comparison applications (ranking configurations, scoring economies). Move-evaluation applications (policy choices, infrastructure commitments, irreversibility detection) do not wait on it.
Contested
reasonable disagreement exists — 12 propositions- Value and Future Option Space contested
Value = configurations that increase future option space for living systems under bounded energy flux.
Where the disagreement livesHow do we measure option space? Over what time horizons? How do we handle trade-offs between different living systems? See option-space-measurability.
- Participation within Physical Limits contested
Economies that maintain long-term social stability tend to decouple basic participation from labour performance.
Where the disagreement livesWhat non-coercive coordination mechanisms scale without falling back on deprivation is a live research and design question.
- Complexity Has a Maintenance Cost contested
Increasing complexity carries an ongoing maintenance cost that grows relative to the benefits it produces.
Where the disagreement livesWhether this pattern applies as a general law or only in specific conditions — Tainter's thesis is contested in historiography — remains open.
- Substitution Has Limits contested
Substitution is empirically bounded by specific physical and ecological invariants that no technology has demonstrated the ability to circumvent. The Daly–Solow–Stiglitz debate (1970s–present) crystallised this as the strong-vs-weak sustainability disagreement; the strong-sustainability position is defended by the empirical specifics (phosphorus, atmospheric sinks, biodiversity, topsoil), not by a general principle.
Where the disagreement livesWhich currently-cited invariants are truly non-substitutable versus not-yet-substituted is a live empirical question — the strong-sustainability case is empirical, not categorical, and remains testable as new technologies arrive. The case is strongest for phosphorus and atmospheric sinks; somewhat softer for individual species (functional redundancy) and topsoil (rebuildable but slow).
- Legibility Trades Off Against Truth contested
Metrics simplify reality at the cost of accuracy; incentives follow metrics; unmeasured factors decay.
Where the disagreement livesWhether better metrics solve this, or whether some values must remain partially illegible to remain what they are, is philosophically live.
- Configuration Is Not Information Alone contested
Configuration is broader than information and cannot be reduced to symbols or data.
Where the disagreement livesWhere exactly the boundary lies between configuration and information, and how to measure the difference, remains under-specified.
- The Viable Objective contested
A viable economic objective is a no-regret preservation criterion over moves: prefer Δω(move) ≥ 0 on R_living(C, B, T) for every major class of life and every relevant horizon T; reject any move that asymptotically empties R_living for some such class at some such horizon. State-comparison is the static frame (partial order over configurations); move-evaluation is the dynamic frame (no-regret over horizons). Both are needed.
Where the disagreement livesThe operational definition of "major class of life" — phylum-level, functional ecological role, capacity for distinction-making? Selection of the relevant horizon for any specific decision (default: the longest horizon any major class depends on, but this is contestable). Gradations between elimination and narrowing of R_living. What to do when no available move avoids long-horizon collapse — probably a fallback like "minimise the horizon at which collapse becomes asymptotic."
-
Labour has functioned historically as an allocation mechanism for access to life-supporting configurations rather than as the source of the value those configurations carry. The conflation between allocation-mechanism and value-source was tractable under historical scarcity regimes (low automation, low energy, limited accumulated structure) because the allocator tracked contribution-to-configuration closely enough for practical purposes. As the composition of scarcity changes, the allocator and the value-source visibly come apart.
Where the disagreement livesWhat proportion of historically-attributed "labour value" was really configuration-contribution being correctly tracked by the allocator, versus rent or coercion being laundered through it, is empirically thick and contested. The marginal-productivity tradition has a serious case in some domains; identifying where that case still holds — and where it has dissolved — is live work, not settled. Whether the analysis transfers cleanly to historically non-wage allocators (kinship obligation, ritual exchange, command economies) or whether each requires its own treatment is open.
- Configuration Generates Configuration contested
For a configuration C with accumulated structure S(C), the reachable set R_living(C, B, T) under a fixed exergy budget B and horizon T strictly contains R_living(C', B, T) for any C' with S(C') ⊂ S(C), where the configurations in S(C) function as substrate for further assembly. The marginal cost of reaching a new configuration C'' from C falls as S(C) grows, because C'' can recruit elements of S(C) as components rather than re-deriving them from the unconfigured state. The proposition is the structural counterpart to asymmetry-of-option-space-change: τ_expand at any single step remains long relative to τ_break, but the number of steps available per unit time grows with S(C), so cumulative expansion compounds where the substrate is preserved.
Where the disagreement livesWhether the rate-of-rate-growth is genuinely structural or is an empirical regularity over specific substrates (biology, post-Enlightenment technology, written mathematics) is open. The strongest articulation available — Walker and Cronin's reading of Assembly Theory as a forward arrow, in which the universe's trajectory through configuration space is what time is — would make the generativity metaphysical rather than empirical. That move is contested in physics and the proposition does not depend on it. The weaker claim — that configurations enable configurations and the rate grows with accumulation — is defensible without the metaphysics. Whether the proposition adds anything to endogenous-growth theory (Romer) beyond physical grounding is also open; the substantive overlap is real but the framing here is configurational rather than knowledge-as-input-to-production.
-
For an economy whose value is generated move by move (configuration-generates-configuration), coordination is the problem of distributed local Δω-evaluation on R_living across multiple observer-relative boards, inside institutions whose role is to widen the set of boards each agent can see (internalising uncounted moves) and to protect the substrate (keeping the reachable set open) — not the problem of discovering a single market-clearing price. Markets are one board within this — fast and useful — not the arbiter.
Where the disagreement livesThe institutions themselves are unbuilt and unspecified — this is a direction, not a mechanism. Coordination across many simultaneous players on shared boards is genuinely harder than one engine on one board, and the chess analogy is weakest exactly there. Whether plural, non-collapsed accounts can resist re-collapse into a single price under real institutional pressure is open. What concretely aims attention at the generative-but-invisible work — care, prevention, maintenance, substrate-building — without metering it into decay is the live design problem.
- The Infinite Game contested
Carse's distinction: a finite game is played for the purpose of winning and terminates in a win; an infinite game is played for the purpose of continuing the play and has no terminal winning state. The no-regret viable objective — preserve R_living(C, B, T) non-empty for every major class of life over every relevant horizon — is the rule of an infinite game: it admits no winning move, only the refusal of game-ending ones. Option space is that game given a physical substrate: R_living is the set of futures in which living systems continue to make distinctions (continue to play), and to preserve it is to keep the game in play. Lineage: James Carse (Finite and Infinite Games, 1986) for the distinction; Buckminster Fuller for the generative means (ever more with ever less, in service of all life).
Where the disagreement livesWhether the identification of the viable objective with Carse's infinite game is a derivation or a productive analogy is itself open — Carse's vision is philosophical and social, and transporting it to a thermodynamic and ecological frame may carry more or less than the original. The hardest edge is the same one viable-objective flags: an infinite game has no rule for what to do when continuation for one class of player requires ending it for another. Naming the telos does not resolve that; it locates it.
- The Political Residue Is a Clock contested
The irreducible political residue of a permission-minimising design is temporal, not spatial — a clock, not a wall: a race against boundary-redrawing, a revocable exemption, or a scheduled opening. The doctrine is to minimise the permission surface and sequence to the domain's clock, gated by the line that routing around rent is never licence to route around protection.
The forming edge
There is no separate drafts queue any more. Live, unsettled thinking enters on the settledness spine at forming — published and not yet load-bearing — and climbs in public; the whole canon's migration is watchable on /motion. What was once filed as a draft has either climbed into canon or remains here, at the open edge, in full.
The candidate formalisation of option space — reachable configurations under a bounded exergy budget, with local Δω as the operational counterpart to the non-computable global measure — is kept in full at the formalisation. Its conclusions already live on the spine as option-space-measurability (open) and option-space-as-chess-moves; the observer-class restriction (R_living) and the four-resolution teaching device that the 2026-04-29 session produced have since landed as canon and as the /polychess surface.
If you're here because you want to engage with the unresolved technical problems, Alexander (in research-frontier mode) is a good thinking partner. Ask about any of the open or contested propositions above.
For the operational counterpart — the rule decisively running on real cases rather than the open questions surrounding it — see /applications/moves. Six worked Δω evaluations across all four verdicts, plus an honest "where this rule runs out" section. That page is where the framework is least open; this page is where it is most.