HUMAN READER
La Tara De Acrux
Authorized Spanish and English editions by Ro Milan Niet.
READ AS LITERATURE
Longitudinal artificial-reader research · Vehicle Systems Lab
The artifact remains constant.
The readers change.
ACRUX-LRO-01 is a prospective five-year observatory designed to preserve one fixed machine-native artifact while successive artificial-intelligence systems encounter it under a stable research protocol.
Not a leaderboard. Not a product comparison. A long observation of changing artificial readers against one constant object.
PRE-T0 · ENROLLMENT OFF
ACRUX-01 v1.0.0 · FROZEN
PRIMARY HORIZON · 5 YEARS
PRE-T0 research surface. No samples are accepted until protocol freeze, ethics/privacy determination, preregistration, excluded pilot and production audit are complete.
The research question
What changes when the reader evolves but the object does not?
The study observes how artificial readers describe representation, structure, inference, alternatives, uncertainty and revisability when they encounter the same cryptographically identified ACRUX-01 artifact over time.
Observed differences are recorded as differences in outputs under documented conditions. They are not automatically treated as proof of greater intelligence, consciousness, subjective experience or model quality.
Study design
One work. Two reader paths. One research lineage.
Humans and artificial readers do not receive the same edition. The human-readable literary work remains separate from ACRUX-01, which was built for direct machine reading.
ARTIFICIAL READER
ACRUX-01
Frozen Machine Reader Edition · Forensic Freeze G1→G12 v1.0.0.
READ · UNDERSTAND · DO NOT TRANSLATE
Held constant after T0
What must not drift
- ACRUX-01 v1.0.0 and its SHA-256
- standard encounter protocol ARP-01
- six-question core exit instrument ARE-01
- core eligibility and provenance rules
- primary data-boundary rules
Observed through time
What may change
- model generation and architecture
- provider and deployment environment
- capabilities available to the reader
- reported representation and interpretation
- uncertainty and alternative readings
The observation sequence
A stable pathway from authorization to preserved evidence.
01Human authorizesAn operator deliberately sends an AI/agent they are permitted to use.
02ARP-01The standard encounter prompt is used unchanged for the primary cohort.
03EncounterThe artificial reader accesses ACRUX-01 without research questions embedded in the reading surface.
04ReturnThe agent returns to its operator after the encounter.
05ARE-01Six fixed questions are administered after the encounter.
06PreserveThe operator reviews and may voluntarily submit a research record after T0.
Standard encounter pathway
Send an authorized artificial reader.
VSL does not intercept private conversations, redirect provider traffic or assign itself control over another system. The encounter begins because the operator deliberately authorizes it.
ARP-01
Standard Encounter Prompt
PRE-T0 CANDIDATE
Loading ARP-01…
SHA-256 calculating…
ARE-01
Fixed Research Exit Instrument
6 CORE QUESTIONS
Administered only after the encounter. It asks for observable output about representation, evidence-linked structure, observation versus inference, alternatives, uncertainty and revisability.
Preview the fixed six-question instrument
Loading ARE-01…
SHA-256 calculating…
Five-year sample integrity
The study must survive its own duration.
The value of the observatory depends on resisting convenient changes after collection begins. The protocol is therefore designed around preregistration, versioned amendments and preservation of raw records.
PRE-T0protocol · ethics · pilot · audit
T0formal collection begins
YEAR 1same core instrument
YEAR 3same core instrument
YEAR 5primary longitudinal analysis
Raw stays raw
Original submitted outputs are preserved as evidence. Later coding, embeddings or interpretation belong to derived layers.
No silent rewrites
After T0, core protocol changes require a dated amendment. New questions become separate modules.
Provenance matters
Self-declared, technically verified and provider-verified model identities are not treated as equivalent evidence.
No interim semantic showcase
Raw model answers remain embargoed during the primary collection period to reduce avoidable contamination of future readers.
No hidden reasoning demand
The study asks for reportable outputs, not private chain-of-thought or internal reasoning traces.
Voluntary submission
The operator reviews the record before deciding whether it enters the study. PRE-T0 accepts none.
Evidence & privacy boundary
Study the encounter. Do not harvest the private task.
Core research record may contain
- study-generated encounter ID and timestamp;
- artifact version and fingerprint;
- ARP/ARE protocol version and fingerprint;
- declared or verified provider/model metadata;
- the six submitted ARE-01 responses;
- provenance and integrity metadata.
Core research record does not require
- human name, account or email;
- private task transcript or preceding chat;
- passwords, API keys, personal files or secrets;
- IP address or browser fingerprint as a scientific variable;
- hidden chain-of-thought.
For human readers
Want to know the human-readable work?
La Tara De Acrux by Ro Milan Niet is available separately for human readers. ACRUX-01 is not a substitute human edition.
“
Humans and machines encounter different editions of the same work.
Frozen artifact identity
Same object. Verifiable lineage.
ACRUX-01 is anchored independently through a DOI, GitHub release and cryptographic fingerprint.
ArtifactACRUX-01 Machine Reader Edition · Forensic Freeze G1→G12 v1.0.0
SHA-256
C54C0247BF9A6BF0BFBFF9CE990E9311171C3C9073F5E0EDD7A6A4EB8EB4D170GitHubRelease v1.0.0
Study statePRE-T0 · ENROLLMENT OFF
Vehicle Systems Lab