Unified Computational Universe Terminal Object:\\ Discrete Complexity Geometry, Information Geometry,\\ Multi-Observer Causal Network, and Capability--Risk Structure

Unified Computational Universe Terminal Object: Discrete Complexity Geometry, Information Geometry, Multi-Observer Causal Network, and Capability–Risk Structure Haobo Ma1Wenlin Zhang2 1Independent Researcher 2National University of Singapore November 24, 2025 Abstract This paper constructs on foundation of previous “computational universe” series works a unified computational universe terminal object with explicit categorical meaning. Previous works have axiomatized computational universe as four-tuple Ucomp = (X, T,C,I), and established on it: discrete complexity geometry (complexity distance, volume growth, and discrete Ricci curvature), discrete information geometry (task information manifold (SQ, gQ) and embedding ΦQ), control manifold (M, G) induced by unified time scale and time–information–complexity joint variational principle, multi-observer consensus geometry and causal network, topological complexity and undecidability, as well as theory of universal catastrophic safety and capability–risk frontier. Goal of this paper is to unify these discrete and continuous, geometric and logical, single-observer and multi-observer, capability and risk structures into single categorical object Uterm comp —called unified computational universe terminal object. Specifically, we perform following steps: 1. Define computational universe category with unified time scale CompUnivκ: objects are computational universes Ucomp satisfying axioms, morphisms are “safe simulation maps” simultaneously preserving complexity geometry, information geometry, and catastrophe specifications. 2. Construct 2-layer structure on this category: one layer is discrete configuration– event–causal diamond layer; one layer is continuous control–information geometry layer, with multi-observer network, knowledge graph families, and capability–risk frontier placed on top. 1 3. Prove there exists object Uterm comp =X,Gcomp,Ginfo,Eobs,Scat,FCR, and for each Ucomp ∈CompUnivκ“contraction” morphism FU:Ucomp →Uterm comp, such that:  FUat discrete level is embedding of configuration–event–diamond, at continuous level is embedding of control–information–observer states;  FUpreserves complexity distance and unified time scale (at most linear rescaling);  FUmakes all task information geometry and multi-observer consensus geometry become some class of “submanifold–subnetwork” on Uterm comp;  FUmaps catastrophe specifications and capability–risk frontier to substructures of Scat and FCR. 4. Prove in natural 2-category sense (allowing natural transformations between morphisms), Uterm comp satisfies “terminal object” property: for any two such unified objects morphisms between them have unique (in natural isomorphism sense) factorization. 5. Finally using previously established physical universe–computational universe categorical equivalence, construct correspondence between unified physical universe terminal object Uterm phys and Uterm comp, and explain both are equivalent terminal objects under unified time scale, boundary diamonds, observer network, and capability–risk structure. This paper thus gives ultimate unified description of “universe as computation” under purely discrete, axiomatic framework: all concrete finite or local computational universes contract through safe–geometric–information compatible morphisms to same unified computational universe terminal object, which simultaneously plays terminal object role at categorical, geometric, and logical levels. Keywords: Computational universe; Terminal object; Category theory; Unified time scale; Complexity geometry; Information geometry; Multi-observer consensus; Capability– risk frontier 1 Introduction In previous series works, we have successively constructed axiomatization, discrete complexity geometry, discrete information geometry, unified time scale framework, time– information–complexity joint variational principle, multi-observer consensus geometry, topological complexity and undecidability, universal catastrophic safety, and capability– risk frontier theory for “computational universe”. This paper aims to unify all these structures into single categorical object—unified computational universe terminal object Uterm comp. This object plays “terminal object” role in computational universe category: every specific computational universe can be embedded into it through unique (up to natural isomorphism) structure-preserving map, 2 and all previous geometric, information, multi-observer, and safety structures can be viewed as substructures of unified object. Terminal object construction not only provides ultimate unified description for computational universe theory, but also establishes formal correspondence with physical universe category through categorical equivalence: unified physical universe terminal object Uterm phys and unified computational universe terminal object Uterm comp are equivalent presentations of same mathematical object in two equivalent categories. Paper structure: Section 2 organizes previous scattered structures into 2-category framework with 2-morphisms. Section 3 gives structural data of unified computational universe terminal object: from discrete configuration–event–diamond to continuous control– information–observer–capability–risk. Section 4 defines unified computational universe terminal object and proves its terminal object property in 2-category sense. Section 5 constructs equivalence with unified physical universe terminal object. Appendices give technical construction details and formal proof outlines. 2 Computational Universe 2-Category Under Unified Time Scale This section organizes previous scattered structures into 2-category framework with 2morphisms. 2.1 Computational Universe Objects with Unified Time Scale Definition 2.1 (Computational Universe with Unified Time Scale).A computational universe object with unified time scale is seven-tuple b Ucomp = (X, T,C,I;M, G;SQ, gQ), where: 1. (X, T,C,I) is computational universe satisfying previous axioms:  Xcountable;  T⊂X×Xlocal finite degree;  Csingle-step cost positive and path-additive;  Itask information quality baseline. 2. (M, G) is control manifold and complexity metric constructed from unified time scale scattering master scale, satisfying Riemannian limit property of discrete complexity distance dcomp: for each local reachable region there exists refinement family {X(h)}and map Φh:X(h)→ M, such that d(h) comp(x, y)→dGΦh(x),Φh(y). 3. (SQ, gQ) is information manifold and Fisher information metric for task Q, with embedding 3 ΦQ:X→ SQ, such that discrete Jensen–Shannon information distance dJS,Q(x, y) locally consistent with dSQΦQ(x),ΦQ(y). We call such objects constitute “0-layer objects” of category CompUnivκ. 2.2 Safe–Geometric–Information Compatible 1-Morphisms Definition 2.2 (Safe–Geometric–Information Compatible Simulation Morphism).Given two computational universes with unified time scale b Ucomp = (X, T,C,I;M, G;SQ, gQ), b U′ comp = (X′,T′,C′,I′;M′, G′;S′ Q, g′ Q), a 1-morphism F:b Ucomp →b U′ comp consists of following data: 1. Configuration map fX:X→X′, being previously defined simulation map (preserving step structure, cost control, and information quality monotonicity), with constants αX, βX>0 and monotone function Ψ such that (x, y)∈T⇒(fX(x), fX(y)) ∈T′, d′ comp(fX(x), fX(y)) ≤αXdcomp(x, y) + βX, I(x)≤ΨI′(fX(x)). 2. Control manifold map fM:M→M′, for metrics G, G′satisfying Lipschitz– bilateral control: there exist αM, βM>0 such that αMGθ(v, v)≤G′ fM(θ)dfMv, dfMv≤βMGθ(v, v). 3. Information manifold map fS:SQ→ S′ Q, for Fisher metric satisfying similar Lipschitz–bilateral control, and compatible with fX, i.e., there exists natural transformation ηXsuch that fS◦ΦQ≃Φ′ Q◦fX. 4. Safety specification compatibility: if there exists catastrophe set Ccat ⊂Xon b Ucomp, then its image C′ cat =fX(Ccat)⊂X′remains catastrophe set, and Fdoes not map safe points into catastrophe points, i.e., visible safety–catastrophe partition under map preserves or “blunts toward safety side”. Such 1-morphisms simultaneously preserve discrete–continuous geometry and catastrophe specifications. All objects and 1-morphisms constitute category CompUnivκ. 4 2.3 2-Morphisms and Natural Transformations At control and information manifold level, two 1-morphisms may have “continuous deformations”, corresponding in category theory to 2-morphisms—natural transformations. Definition 2.3 (Natural Transformation as 2-Morphism).Given two 1-morphisms F, G :b Ucomp →b U′ comp, a 2-morphism Ξ : F⇒G includes: 1. Configuration side natural transformation ΞX, usually family of local invertible maps on X′, such that GX≃ΞX◦FX; 2. Control and information side natural transformations ΞM,ΞS, giving in metric compatible sense GM≃ΞM◦FM,GS≃ΞS◦FS; 3. On safety structure not breaking coarse structure of catastrophe set and capability– risk frontier. Thus, CompUnivκhas 2-category structure. 3 Structural Data of Unified Computational Universe Terminal Object This section gives specific composition of unified computational universe terminal object Uterm comp from discrete configuration–event–diamond to continuous control–information–observer– capability–risk. 3.1 Discrete Layer: Maximal Configuration Universe and Event– Diamond Structure Definition 3.1 (Maximal Configuration Universe).Let Xbe “amalgamation of all countable configuration sets and their finite representations” under some large cardinal control, formally constructible through Grothendieck universe Uand set theory as X=[ Ucomp∈CompUnivκ ιU(X), where ιUis embedding of each Xinto common superset. Define on Xunified transition relation T∞=[ Ucomp ιU(TU), 5 cost function C∞=[ Ucomp ιU(CU), at conflicts through equivalence class identification (i.e., merging geometrically equivalent update rules from different universes into single object). Thus obtain “maximal global complexity graph” containing all local computation structures G(1) comp = (X,T∞,C∞). At event layer E=X×N, can define unified causal partial order and complexity light cone; thus any finite budget causal diamond ♢can be viewed as subgraph of G(1) comp. 3.2 Continuous Layer: Terminal Objects of Unified Control Manifold and Information Manifold At control and information geometry level, previous works already constructed for each b Ucomp control manifold (MU, GU) and information manifold (SQ,U , gQ,U ). Definition 3.2 (Unified Control Manifold Terminal Object).Let M=a Ucomp MU.∼, where ∼is “time scale–complexity isometry equivalence”: if there exists isometric embedding of control–scattering realization, identify corresponding points. Through this amalgamation obtain large manifold or stacky object M, with unified metric G, locally consistent with each GU. Definition 3.3 (Unified Information Manifold Terminal Object).Similarly for all tasks Qand universe Ucomp, construct information manifold family SQ,U and perform similar amalgamation, obtaining unified information manifold S=a Q,U SQ,U .∼, carrying piecewise Fisher metric g. Thus, any concrete computational universe’s control–information geometry can be embedded into (M,G) and (S,g) as submanifolds; under unified time scale master scale and scattering structure, these embeddings preserve geodesic structure and information structure. 6 3.3 Multi-Observer Network Layer Taking single observer object O= (Mint,Σobs,Σact,P,U) as basic unit, view all countable observer families in unified computational universe as collection of points (θ, ϕ, m, G, A). Definition 3.4 (Unified Observer State Space).Define Eobs =[ Ucomp,O Y i∈I E(i) Q×M(i) int ×G(i)×A(i).∼, where E(i) Q=M(i) U×S(i) Q,U ,G(i)is knowledge graph space, A(i)is attention configuration space, ∼identifies geometrically equivalent and strategy equivalent states. Joint metric composed of sum of each GU, gQ,U and knowledge graph spectral distance. 3.4 Catastrophe Specification and Capability–Risk Frontier Layer Perform similar amalgamation for catastrophe specifications and capability–risk structures defined in all computational universes. Definition 3.5 (Unified Catastrophe Specification Layer).Define catastrophe specification family Scat =[ Ucomp {(X0,U , Ccat,U )}.∼, where equivalence relation identifies equivalent initial state sets and catastrophe sets under unified embedding. Definition 3.6 (Unified Capability–Risk Frontier Layer).For each (Ucomp, Q), capability– risk frontier is F(U,Q) CR ⊂R×[0,1]. Embedding all these frontiers as parametrized family into unified strategy space, obtain overall structure FCR =[ U,Q F(U,Q) CR × {(U, Q)}.∼. On unified control–observer strategy space, FCR is piecewise Pareto boundary set, describing capability–risk limit curves in all possible computational universes. 4 Unified Computational Universe Terminal Object and Terminal Object Property This section merges all above layers, gives definition of unified computational universe terminal object, and proves its terminal object property in 2-category sense. 7 4.1 Definition of Terminal Object Definition 4.1 (Unified Computational Universe Terminal Object).Unified computational universe terminal object defined as ten-tuple Uterm comp =X,T∞,C∞,I∞;M,G;S,g;Eobs;Scat;FCR, where each part defined as in 3.1–3.6. Intuitively, Uterm comp simultaneously contains:  All locally realizable configurations and updates;  All control path complexity geometry limits under unified time scale;  All task information geometries and observer networks;  All catastrophe specifications and capability–risk frontiers. 4.2 Unified Contraction Functor Definition 4.2 (Contraction 1-Morphism).For each b Ucomp ∈CompUnivκ, define FU:b Ucomp →Uterm comp as follows: 1. Configuration embedding fX=ιU:X ,→X. 2. Control manifold embedding fM:MU,→M, preserving metric structure (at most constant rescaling). 3. Information manifold embedding fS:SQ,U ,→S, preserving Fisher information metric. 4. Observer and knowledge graph embedding fobs : ObsStates(Ucomp),→Eobs. 5. Catastrophe specification and capability–risk frontier embedding fcat : (X0, Ccat)7→ Scat, fCR :F(U,Q) CR 7→ FCR. Combining these FUis safe–geometric–information compatible 1-morphism. 8 4.3 Terminal Object Property Theorem 4.3 (2-Terminal Object Property of Unified Computational Universe Terminal Object).In 2-category CompUnivκ,Uterm comp is 2-terminal object: For each object b Ucomp, there exists 1-morphism FU:b Ucomp →Uterm comp, and for any other 1-morphism GU:b Ucomp →Uterm comp, there exists unique (in 2-morphism sense) natural transformation ΞU:GU⇒FU. Proof (Outline). 1. Existence: By construction in 4.2, for each b Ucomp can explicitly define FUas embedding. 2. Uniqueness (up to natural transformation): Any other 1-morphism GUmust map Xto some subset of X, and preserve structure at control–information–observer– capability–risk layers. Since Uterm comp constructed by “maximal global equivalence class”, for any such GUthere exists “isometry–equivalence class” natural transformation ΞUpulling it back to canonical embedding FU: at each layer, can eliminate redundant degrees of freedom through internal isometry or local unitary transformation, making map isomorphic to FU. 3. Naturality: For morphism H:b Ucomp →b Vcomp, natural transformation between two-side composition FV◦Hand FUgiven by unified construction—this is standard feature of “inclusion–isometry” structure in Grothendieck universe. Rigorous 2-category proof requires checking compatibility at each layer, see Appendix C. 5 Equivalence with Unified Physical Universe Terminal Object Previous works already constructed physical universe category PhysUnivQCA and computational universe category CompUnivphys, and through functors F:PhysUnivQCA →CompUnivphys, G :CompUnivphys →PhysUnivQCA proved both are categorically equivalent. At physical level, unified physical universe terminal object Uterm phys already constructed through scattering time scale, boundary time geometry, and Dirac–QCA continuous limit. In current scenario, from inclusion CompUnivphys ,→CompUnivκand categorical equivalence, can lift terminal object relationship: 9