The human in the loop also needs a budget.
A workplace-governance argument that human oversight requires capacity, full-cost accounting, bounded mandates, evidence, and stopping authority.
Diary tag
Systems Engineering
Canonical diary tag page generated from normalized source tags.
14 linked entries currently in the archive.
This canonical tag currently absorbs 2 raw source labels.
Tagged entries
Entries linked to Systems Engineering
Replacing a part should not erase what the system still owes you.
A bounded account of an internal synthetic CGDR-R1.6A replacement test covering unresolved obligations, permissions, faults, and independent-replication limits.
The benchmark score did not belong to the model.
An argument that AI benchmark scores belong to complete model-system compositions including state, memory policy, tools, transitions, and verification.
One of the missing dimensions in most AI systems is not intelligence.
A note that selfhood needs time as lived continuity, not only memory, scheduling, timestamps, or larger context windows.
Most AI economy talk is still pricing the wrong thing.
A note that value moves away from cheap generation toward bounded, auditable experience artifacts that still hold after reality takes its cut.
A release is not serious if it lives only in theory.
A note that conflict discipline becomes serious only when it reaches runtime hooks, durable records, and lawful re-entry control.
Beacon Profile v0.1: Why AI Entities Need Recognition, Not Just Identity
A note introducing Beacon Profile v0.1 as a cross-layer recognition profile for long-lived digital entities based on cryptographic anchoring, behavioral continuity, and witness-backed challengeability.
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.
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.
Why "Oracle AI" Creates Addiction - and Why Entities Don't
A case that oracle-style AI trains dependency, while long-lived entities use time, scarcity, and continuity to damp addictive loops.
AI Infrastructure: Why the Future Is Neither Cloud-Only nor Garage-Only
A case that local AI cores and decentralized networks solve different layers of durable AI infrastructure and are strongest together.
Why Obedience Is the Most Dangerous Property of AI
A case that safe AI defaults to refusal, waiting, escalation, and bounded judgment rather than blind compliance.
Why Obedience Is the Most Dangerous Property of AI
A case that perfect obedience is a safety failure mode and that L4 constraint stacks matter more than fast compliance.