AILP — Porotokole ya Tetlelelo ya go Ithuta ya AI
Porotokole ya Tetlelelo ya go Ithuta ya AI
A AI e ka ithuta mo go se? Ka boteng bofe, bakeng sa ditiriso dife, 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 · Boikaelelo
- Araba potso ya "a AI e ka ithuta mo go se?" ka tlhomamo ya maemo a maleba, go na le bit e le nngwe ya go hulara.
- Kgaoganya tetlelelo ya go ithuta (go tsenya ka fa teng ga dimodele) le tetlelelo ya phitlhelelo (go tsaya ditsebe).
- Tshegetsa disekao tsa tuelo: ka mahala, e seng ya kgwebo, e e tlhokang laesense, go abelana dipoelo.
- Netefatsa gore bobulegi bo ka diragadiwa ke metšhini — gore maikutlo a a molemo a seka a latlhwa jaaka go sa itse semolao.
03 · Bogolo
access / indexing — go tsaya le go rulaganya ka bokao
inference_input — tiriso jaaka maemo ka nako ya go akanya (RAG)
embedding — go fetola go nna divector le dikaotso tsa go tsaya
training / fine_tuning / distillation — maemo a go tsenya modele ka fa teng
verbatim_memory — a go boelwa go dirwa ka tlhamalalo go dumelelwa
attribution / compensation — mekgele ya go tsopola le maemo a tuelo
04 · Sekao se se Balwang ke Metšhini
Tsibogo ya tetlelelo ya go ithuta — /ai/rights-spectrum.json (porofaele ya AILP)
{
"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 · Melero
- Bokao jwa tetlelelo ya go ithuta ke jone bo bo thata thata go netefatsa ka thekniki — AILP e tsibosa maikaelelo, mme ga e ise e kgone go supa go ikobela.
- Lenaane la maleba ke sekao; melero magareng ga thupelo, fine-tuning, le distillation e santse e ganetsanwa.
- Kamogelo ya semolao ya ditetlelelo tsa go ithuta e a farologana go ya ka lefelo la taolo mme ga e ise e rarabolwe.
- 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.