What is the ARC Principle?
The ARC Principle (ARC: Artificial Recursive Creation) is Law I of Michael Darius Eastwood's ARC Theory, a law under test written U = I x R^alpha: usable capability U equals base intelligence I times recursive depth R raised to an exponent alpha. He first stated it on 8 December 2024. The exponent is measured, never assumed.
Text reviewed 2026-10-31. Any earlier audio edition may differ; the dated text and linked records govern.
His sealed record of 8 December 2024 states it as the ARC Hypothesis, in the form U = I x R. Infinite Architects printed it on 2 January 2026 with the exponent fixed at two. Paper I (17 January 2026, DOI 10.17605/OSF.IO/B6N27) formalised it with the exponent left open and first tested it. Each term has an operational definition: R counts completed improvement passes on a ladder fixed in advance, and U and I are scores on a battery frozen before the first measurement.
The early single-model estimate of about 2.24 was retracted when a blinded replication on five frontier models did not reproduce it. Paper II reports error-decay estimates from one model: about 0.49 by regression (standard error 0.20) and 0.59 by endpoint calculation, whose bootstrap interval runs from -1.3 to 2.9. The result is inconclusive. These do not directly measure a capability-growth exponent or self-referential coupling.
Recursion as a structural idea has long prior work, in evolutionary theory, self-modifying computation and quantum error correction; the priority claimed is for this named relation and its cross-domain framing, dated 8 December 2024. It is distinct from Law III, the ARC Ceiling, which is registered as P20, with no verdict yet: a conditional specialisation of Law II, derived inside a pacing model, so a minimal-model result rather than a general law.