A small external comment turned into a useful control-layer correction.
Announcement of the Self-Evo WDC v0.1.2 release-candidate package on Witness Dependency Capture and the survivability of independent witnesses.
Diary tag
Human Oversight
Canonical diary tag page generated from normalized source tags.
11 linked entries currently in the archive.
This canonical tag currently absorbs 2 raw source labels.
Tagged entries
Entries linked to Human Oversight
Humanoid robotics shows that AI safety is becoming operational and physical.
A note that embodied AI safety becomes a property of the full operating environment, including bodies, homes, privacy, evidence, and L4 consequences.
L4 in Practice: 5 Reality Signals That Kill "Smart" Systems
A note that cost, heat, time, maintenance, and human bandwidth are the signals that determine whether long-lived AI survives contact with physics.
Visual Experience Capsules (VXCX) - Why "what you see" matters more than pixels
A note introducing VXCX v0.1 as an L2 protocol for sharing visual experience capsules without transmitting raw pixels by default.
The EU AI Act is landing in the real world - and the timing is not accidental.
A note that the EU AI Act is arriving as a compliance timeline and evidence discipline, with embodied systems making responsibility procedural.
Ester Clean Code - v0.2.1 is out (with v0.2.0 as the hardening baseline).
A release note for Ester Clean Code v0.2.1 that frames hygiene, fail-closed defaults, and auditability as the basis for long-lived local-first systems.
HGI - why am I only hearing this now? And why "General" is never a default.
A case that HGI is an overloaded acronym and that claims about "general" intelligence need an explicit reference class, human anchor, and audit trail.
L4 in Practice: 5 Reality Signals That Kill "Smart" Systems
A case that cost, heat, time, maintenance, and human bandwidth are the real signals that determine whether long-lived AI survives contact with physics.
From Better Chat to Stable Presence
A case that stable agent presence requires continuity, constraints, and durable audit trails rather than better chat alone.
An AI paid a human to hold a sign. Funny - and also a boundary crossing.
A case that AI-mediated physical action becomes safe only with verified identity, hard budgets, human vetoes, and durable witness trails.
Ads in private AI chats are not the future. Utility is.
A case that private AI should deliver consent-first utility and audited recommendations rather than ad-based chat.