sugarlicks's avatar
sugarlicks
sugarlicks@verified-nostr.com
npub13zhj...vg5k
creative technologist + lawyer | arts, culture producer | Web3 / DAO + identity | Lebanese Kiwi #tangatatiriti | siltcore.org
sugarlicks's avatar
garethfakhry 3 days ago
The internet is increasingly becoming an institutional web, with coordinated auth across systems ... but no answer to the Q: By what authority does this action occur? Identity systems solved authentication. They rarely solved legitimacy. New essay:
sugarlicks's avatar
garethfakhry 2 weeks ago
Authority does not appear spontaneously inside systems. Every legitimate act carries a lineage: source, mandate, responsibility, consequence. Computational systems increasingly execute action without carrying the semantic conditions surrounding authority itself. The result is a form of rootlessness now emerging across digital infrastructure and AI systems. My latest essay explores this fracture and the need for semantic legitimacy structures capable of preserving rooted authority within machine environments. “The Problem of Rootlessness”
sugarlicks's avatar
garethfakhry 3 weeks ago
AI agents can transact, negotiate, allocate capital, invoke tools, and increasingly act within the physical world itself. The deeper question is no longer identity alone. It is authority lineage. By what authority does an action bind? What mandate travels with it? Who carries continuity, responsibility, and revocation as authority moves through systems? New essay: “The Re-emergence of Authority Lineage” #SILT #AuthorityLineage #AI
sugarlicks's avatar
garethfakhry 3 weeks ago
Consent online isn’t consent. It’s a timestamp. AI agents don’t understand consent. They inherit it. Old permissions look the same as valid ones. No context. No expiry. No boundary. So they act. And the system calls it authorised. The problem isn’t the agent. It’s the missing structure (of consent)
sugarlicks's avatar
garethfakhry 1 month ago
Most digital systems feel complete. They can identify who is acting, verify what that actor holds, and determine what actions are permitted. And yet something still breaks. An action can pass every check and still be illegitimate. Because systems model authorisation, not authority. They can tell you whether an action may proceed, not whether it is justified. That distinction has always existed in law and institutions. We’ve simply failed to encode it into our systems. This is the gap. New writing >>
sugarlicks's avatar
garethfakhry 1 month ago
Even if identity is self-sovereign, authority is still relational. As is human action and transaction. And most systems don’t model it properly. They verify identity. They don’t verify capacity to act. That’s where things break. v0.1 of SILT Core →
sugarlicks's avatar
garethfakhry 2 months ago
Authentication proves control of a key. It does not encode authority. An AI agent told to “explore its capabilities” created a dating profile on behalf of its user. The system worked exactly as designed. The missing layer is mandate architecture. New essay: