Edition status. Scientific reconstruction, version 4.1, 21 October 2026: a working draft, released before its review is complete. Version 4.1 corrects three statements against version 1.103 of the theory registration (15 October 2026): Laws IV and V are registered as laws and untested, Law II is quoted as version 1.103 states it at the boundary, and the mathematics of Laws IV and V is cited at version 1.1. Open review. Still to run: review from outside the programme by a statistician, by a cosmologist or quantum-gravity specialist and by a philosopher of probability, without which this version is not final; a systematic literature review of the priority of the exact conjunction claimed in Section 14.5; and the deposit, beside this paper, of the reference code described in Appendix B. When they have run, the checked paper replaces this one as a labelled new version, with its changes listed.
(version 4.1) · OSF record · DOI 10.17605/OSF.IO/4PGVC
Theory-level predictions preregistered on OSF: DOI 10.17605/OSF.IO/P8CKQ, registered 8 September 2026 (version 1.100), updated 13 September 2026 (version 1.102), updated 15 October 2026 (version 1.103), twenty-two propositions.
Companion: Register · The ARC Theory · Operational Definitions, version 1.9.1, 5 October 2026.
Other fields
Recursive Dynamics · The ARC Theory · Recursive Creation, the hypothesis
Could intelligence create real universes, and what would count as evidence?
Working Paper · The Hyperspace Recursive Intelligence Hypothesis: a Bayesian and Physical Research Programme for Intelligent Cosmogenesis
The scientific core of HRIH, whose long-form edition is the formal statement of Recursive Creation, a speculative hypothesis reachable from Recursive Dynamics but never contained by it: none of Laws I to V is a premise of it or enters any likelihood in it, the ARC Theory loses nothing if it is refused, ARC work may inform its creator-stability factor, and neither the Eden Protocol nor the HARI Treaty (Hardware-Aligned Recursive Intelligence) enters it. Its other neighbours are the statement paper and the founding paper of Recursive Dynamics.
Abstract
To the question of why a universe has the laws, constants and boundary data it has, the Hyperspace Recursive Intelligence Hypothesis (HRIH) answers with intelligent cosmogenesis. This paper reconstructs it as a nested, falsifiable research programme, not as a claim that intelligence has been shown to create universes. It distinguishes (i) physically real intelligent cosmogenesis, in which an advanced agent instantiates a daughter cosmology; (ii) recursive cosmogenic lineages, in which such events occur over more than one generation; and (iii) a stronger closed-cycle submodel, postulated rather than derived from recursion, in which the whole history is a globally self-consistent fixed point or periodic orbit, so that the Big Bang need not be an unexplained first event. The paper specifies a fair model space containing non-deliberate cosmology, simulation families and real intelligent cosmogenesis, and derives that naive observer counting returns the analyst's prior rather than a fact about the world; it requires explicit observer weighting, defines a cosmogenic reproduction number and states the extinction threshold for a branching lineage. Its principal discriminator, a hypothesis for held-out tests to decide, is maker-selection beyond habitability: weighting laws and constants by independently defined cosmogenic capability, stable technological ascent and viable daughter yield. The one observation already in hand, the order of the initial condition, is used only in conditional form, and five predictions and five studies are specified for registration, each with what would count against it. The reconstruction withdraws seven invalid inferences: recursion does not abolish boundary conditions; occurrence does not imply constructibility; ability does not imply preference for real cosmogenesis; broad recursive analogy is not support; unregistered probability ranges are not posteriors; lower-level ARC experiments do not test the cosmological ontology; and the closed loop is a nested stronger submodel rather than the premise of all intelligent cosmogenesis. Current evidence is non-discriminating between HRIH and its strongest rivals, and this paper assigns HRIH no posterior probability; each central proposition can acquire or lose probability separately.
Michael Darius Eastwood conceived and directs this research programme and is the author of this work. Across the programme, he has used more than six AI systems in parallel, under his own instructions, to stress-test his arguments, identify possible errors, and assist in preparing draft text from his own outlines. He determines what is adopted, revised or rejected and takes responsibility for the published content. These systems are tools, not authors.