Uphando

Amalungelo, imisebenzi, uxanduva, kunye nolawulo lwe-AI, i-AGI, kunye nee-arhente — ukusuka kwizixhobo, ukuya kubasebenzi-ndawonye, ukuya kwizihlokwana ezinokubakho kwixesha elizayo. Oku kuquka umbuzo wokhuseleko olusezantsi lobuqu: ziziphi imigqaliselo efanele ilawule usebenzelwano lomntu ne-AI ngaphambi kokuba imibuzo yobuntu isonjululwe.

What AI systems may do with content: read, summarize, transform, retrieve, train, commercialize, redistribute. The AICR ruleset and the AICL-C content-licensing layer make these rights declarable and transactable.

Enoba i-AI inokufunda, kubunzulu obunjani, ngeenjongo ezinjani, kunye nangaphantsi kweembophelelo ezinjani. Ukufundwa akukho njengokufundelwa kuko; 'ukungathintelwa' akukho njengoko kuthiwa 'kunokufundwa.' I-AIRS kunye ne-AILP ziguqula imvume yokufunda ibe luluhlu olunyukayo, olufundwa yimatshini.

Indlela ii-arhente ezifikelela ngayo kwiiwebhusayithi, ii-API, iidatabheyisi, izikhondlo zolwazi, kunye nomxholo ohlawulelwayo: isazisi, imvume, izicelo, intlawulo, ugunyaziso, iirekhodi zokusetyenziswa, imida yokhuseleko, kunye nokuzikhusela kuhlaselo lwe-prompt-injection.

What an Agent may DO once it holds tool or API capability, as distinct from what it may read or learn: action taxonomy, effect vectors, reversibility, blast radius, and multi-action composition risk. AARS gives content permission and action permission separate, independently-evaluable axes.

Who an Agent acts for, where its authority came from, whether it can hand that authority to another Agent, and how much of its internal state a verifier may demand. AADP separates principal from actor, bounds delegation to strictly narrow (never silently widen), and treats inspection as a bounded right, not an unlimited one.

Imigaqo yolawulo enokufunyanwa nokusetyenziswa yimatshini: i-llms.txt, iimanifesti ze-/ai/, iifayile zomgaqo ze-/.well-known/, iiSkim ze-JSON, iitokheni zelayisensi ezisayinwe, kunye neefomathi zeerekhodi zohlolo-mfihlo. Le sayithi ngokwayo ngumzekelo osebenzayo.

Uqoqosho lwezopolitiko lwe-AI nomxholo: ukuhlawula ngokufunyaniswa (pay-per-crawl), izabelo zenzuzo yedatha, iifondi ze-AI zolawulo, iiphuli zembuyekezo yabadali, iimarike zedatha enamanqanaba, kunye nendlela ixabiso elikhutshwe kulwazi lwasekuhlaleni elinokubuyela kwabo balivelisayo.