I did not begin this work by trying to invent a new unit of intelligence.
I began by trying to build a machine whose internal claims could be separated from evidence, whose ideas could be separated from their physical implementations, whose failures could remain part of its history, and whose improvements could be tested against reality before they became authoritative.
That work became AIEN, Omega, PHYSICS, FORGE, AEGIS, ARGUS, Cortex, and the resident reaction architecture now implemented across the AIEN repositories.
While working inside that layer of abstraction, I encountered a measurement problem: the system could generate alternatives, verify them, measure them on real hardware, preserve failed hypotheses, discover recurring abstractions, promote verified improvements, and survive injected failures — but there was no principled way to say how much understanding any of it constituted. The Turing is the answer this engineering work forced.
What this paper is
This is the engineering companion to the formal preprint “From the Imitation Game to the Turing”. It documents the architecture and the evidence program — AIEN, Omega, the sovereign-lineage repositories, the public execution receipts — from which the measure emerged. It makes no independent formal claims: definitions, proofs, the measurement protocol, and falsification criteria live in the formal paper, which is the reference document.
The architecture in one picture
The stack the measure emerged from runs on a discipline of separation: the part that proposes is not the part that verifies, the part that verifies is not the part that authorizes, and every step leaves evidence behind. The map below restates the companion paper's architecture figure for the web.
Figure · The architectureConceptual schematic
The architecture the measure emerged from. Generation flows in one direction: AIEN proposes, Omega defines, synthesizes, and verifies, and FORGE realizes. AEGIS and ARGUS sit outside that path, verifying and observing without deciding or authorizing. Evidence records everything, and the dashed line on the left is the loop back: receipts feed the next round. Conceptual schematic, after the companion paper's architecture figure.
The loop the architecture runs
Inside that separation, every claim travels the same loop. Weights suggest candidates. Programs carry the meaning in a form that can be checked. Omega verifies. Physics authorizes what actually runs, on real hardware at real cost. And the evidence left behind teaches the next round.
Figure · The doctrine loopConceptual schematic
The doctrine loop. Model weights only suggest. The meaning lives in programs, Omega verifies them, physics authorizes what actually runs, and the recorded evidence teaches the next round. No step can be skipped, and no step may do another step's job. Conceptual schematic.
Research artifacts
Companion paper (PDF)
25 pages. The engineering history: the AIEN/Omega architecture, the public evidence receipts, and the first experimental proposals.
@misc{stapleton2026turingemergence,
author = {M. Drake Stapleton},
title = {Computing Machinery and Understanding:
The Turing as a Measure Emerging from AIEN},
year = {2026},
month = {September},
version = {1.9},
note = {Preprint, not peer reviewed.
Companion to "From the Imitation Game to the Turing"},
url = {https://aienos.com/research/computing-machinery-and-understanding-emergence}
}