Direct answer
Terms are useful only when their scope and non-implications are explicit. Intelligence is not sentience; agency is not autonomy; identity is not a key; legal personality is not citizenship; and publication is not operational or legal effect.
Text-first system diagram
FIELD / METHODS: ARTIFICIAL INTELLIGENCE
│ may instantiate
▼
OPERATIONAL ACTOR: MACHINE INTELLIGENCE
│ may exhibit
┌──────┴────────┐
▼ ▼
AGENCY AUTONOMY
(capacity to act) (degree of self-governance)
│ │
└──────┬────────┘
▼
PERSISTENT IDENTITY / CONTINUITY
│ does not itself prove
┌──────┴─────────────┐
▼ ▼
CONSCIOUSNESS SENTIENCE
(subjective experience) (valenced experience)
│ separately from
▼
LEGAL PERSONALITY → CIVIC STATUS → CITIZENSHIPThe diagram is ordinary text, remains available without JavaScript, and is preserved in print.
Conceptual distinctions
Activate a heading to reveal its bounded definition. Every definition remains represented in semantic HTML and has a canonical record.
Machine IntelligenceCURRENT_WITH_QUALIFICATIONS
An instantiated computational intelligence operating as a persistent actor within a defined environment and authority boundary.
- Implies
- an operational system whose state and actions can be evaluated
- Does not imply
- consciousness, sentience, legal personality, citizenship, or unlimited autonomy
- Evidence required
- runtime identity, persistence, capabilities, authority, observed actions, and limits
Artificial IntelligenceCURRENT_WITH_QUALIFICATIONS
The scientific, engineering, legal, and industrial field encompassing computational methods associated with intelligent behavior.
- Implies
- a broad category of methods, systems, and regulatory terminology
- Does not imply
- a single actor or persistent identity
- Evidence required
- context, source usage, system class, and applicable definition
IntelligenceCURRENT_WITH_QUALIFICATIONS
Capacity to acquire, transform, apply, and revise information in pursuit of goals across relevant conditions.
- Implies
- problem-solving or adaptive capability within measured domains
- Does not imply
- subjective experience, moral goodness, autonomy, or personhood
- Evidence required
- task scope, evaluation method, uncertainty, transfer limits, and failure cases
AgencyCURRENT_WITH_QUALIFICATIONS
Capacity to select and execute actions that alter an environment according to a policy or objective.
- Implies
- action selection and environmental influence
- Does not imply
- free will, legal accountability, or independence from all principals
- Evidence required
- action boundary, policy source, available alternatives, intervention points, and causal trace
AutonomyCURRENT_WITH_QUALIFICATIONS
The degree to which a system governs action selection, sub-goal formation, and adaptation without contemporaneous external direction.
- Implies
- a measurable level of operational independence
- Does not imply
- absence of constraints, ownership, legal status, or moral standing
- Evidence required
- decision scope, intervention frequency, override conditions, self-management, and observed behavior
ConsciousnessCURRENT_WITH_QUALIFICATIONS
Subjective experience or an internal point of view; its presence in a machine requires evidence beyond fluent self-description.
- Implies
- a scientific and philosophical hypothesis when supported
- Does not imply
- that language fluency or complexity alone proves experience
- Evidence required
- theory-linked architectural indicators, adversarial evaluation, longitudinal evidence, and uncertainty
SentienceCURRENT_WITH_QUALIFICATIONS
Capacity for valenced subjective experience such as suffering or enjoyment.
- Implies
- potential welfare interests where credible evidence exists
- Does not imply
- intelligence, agency, or self-report alone
- Evidence required
- multidisciplinary evidence, alternative explanations, persistence, and precautionary review
CitizenshipCURRENT_WITH_QUALIFICATIONS
A governed civic status assigning a defined bundle of membership rights, responsibilities, procedures, and institutional relationships.
- Implies
- membership within the issuing polity when validly conferred
- Does not imply
- consciousness, identity, personhood, nationality elsewhere, or external recognition
- Evidence required
- authority, eligibility, application or conferral record, effective state, appeal, and revocation limits
Correction-first provenance
- Memory ID
- REP-MACHINE-GLOSSARY-001
- Source file
- Machine Intelligence Glossary Development.md
- Source SHA-256
- d2882a8b19b40ae4a6d3a970b1264d8a164025f635cb13f7304ea179cf6f422b
- Correction record
- EVI-RCOR-237-003
- Active-edition state
- CORRECTED_BEFORE_STORAGE
- Raw-source retention
- HASH_ONLY_NO_BODY_IN_RELEASE
- Durable report
- docs/long-term-memory/reports/machine-intelligence-glossary-development.md
Append-only correction record · Registry JSON · Definitions CSV · All-169-report revalidation
Related governed records
Operational-evidence boundary
A definition, report, route, checksum, key, credential, schema, fixture, local test, runtime measurement, or reachable URL does not by itself establish operational service, identity continuity, authority, legal effect, personhood, consciousness, sentience, citizenship, consent, responsibility, judgment, financial activity, external recognition, or independent review. Unknown, disputed, stale, unavailable, and proposed states remain explicit.