AILP — Phrothokholo ya Tumelelo ya go Ithuta ya AI

Phrothokholo ya Tumelelo ya go Ithuta ya AI

Na AI e ka ithuta se? Go ya botebong bofe, bakeng sa merero yefe, ka maikarabelo afe?

01 · Tlhaloso

AILP operationalizes the AIRS spectrum for the specific question of learning. It distinguishes acts that binary crawler rules collapse together: being read is not being learned from; being retrieved is not being trained on; fine-tuning is not distillation; verbatim memorization is different from statistical internalization. AILP lets a publisher declare, in a single machine-readable file, per-dimension permissions — access, indexing, inference input, embedding, training, fine-tuning, distillation, memory, output, attribution and compensation — each as allowed, denied, or license-required. An AILP grant answers "may AI learn from this content?" only — it does not answer "may this Agent invoke a tool, write to a database, or publish?" (AARS), nor "who authorized this Agent?" (AADP).

02 · Morero

Araba potšišo ya "na AI e ka ithuta se?" ka go nepagala ga dintlha tše di arogilego go e na le bit ya crawl fela.
Arola tumelelo ya go ithuta (go tsenya ka gare ga dimmotlolo) go tumelelo ya phihlelelo (go hwetša matlakala).
Thekga mehlala ya moputso: mahala, ye e sego ya kgwebo, ye e nyakago laesense, go abelana poelo.
Netefatša gore bulego bo ka phethagatšwa ke metšhene — gore boikemišetšo bjo bobotse bo se lahlelwe thoko e le go se kgonthiše ga semolao.

03 · Tekanyo

access / indexing — go tsopola le go hlopha ka moko

inference_input — tirišo bjalo ka konteks nakong ya go nagana (RAG)

embedding — vectorization le dinamane tša go tsopola

training / fine_tuning / distillation — maemo a go tsenywa ga mmotlolo

verbatim_memory — gore na go hlangwa gape ka mo go nepagetšego go a dumelelwa

attribution / compensation — maikarabelo a tsopolo le maemo a tefo

04 · Mohlala o balwago ke metšhene

Phatlalatšo ya tumelelo ya go ithuta — /ai/rights-spectrum.json (profaele ya AILP)

/ai/rights-spectrum.json

{
  "version": "0.1",
  "protocol": "AILP",
  "publisher": "example.org",
  "default": {
    "access": "allowed",
    "indexing": "allowed",
    "inference_input": "allowed",
    "embedding": "allowed",
    "training": "license_required",
    "fine_tuning": "license_required",
    "distillation": "denied",
    "verbatim_memory": "denied",
    "attribution": "required",
    "compensation": "contact"
  },
  "contact": "licensing@example.org"
}

05 · Mellwane

Moko wa tumelelo ya go ithuta ke wona o boima kudu go o netefatša ka thekniki — AILP e phatlalatša maikemišetšo, eupša ga e sa kgona go bontšha go latela melao.
Lenaneo la dintlha ke moralo; mellwane magareng ga thupelo, fine-tuning, le distillation e sa dutše e ngangišanwa.
Tsebagalo ya semolao ya ditumelelo tša go ithuta e a fapafapana go ya ka lefelo la semolao gomme ga e sešo ya rarollwa.
v0.1.1: AILP(Resource,ContentUse)=Allow does not imply AARS(Actor,Action)=Allow, and does not imply AADP(Principal,Actor,Authority)=Valid — a system exposing both content and tools must evaluate them separately.