Epictetus
5224645703
- docs/vision/LEONARD_FEEDBACK_TO_ARCHITECT.md: 4 resolutions (naming aiND vs Athena; T6 practitioner-gate; T1 has no engine signal yet/actionability_score reserved; ingestion audit is mandatory for indispensable editions) + 2 risks (success metric is human-only; schema forward-compatible, no change in 1.5). - docs/vision/AI_NEWS_DAILY_TIERS.md: 7-tier hierarchy + Builder-Outcome DNA + 4 editorial questions + T6 gate. LOCKED reference for edition curation. - Issue_002.md: PROPOSED hand-curated edition from live 60-story window; T1-T5 + T6 gate; founder confirmation required before 'shipped'. ~11/60 cleared Editorial Test -> exposes ingestion-supply problem. Sprint 1.5 = multiple hand editions; Lens/Memory/Narratives/KG deferred. No schema change.
2.5 KiB
AI News Daily — Story Tiers & Builder-Outcome DNA
Publication: aiND (see AI_NEWS_DAILY_VISION_V2.md)
Status: LOCKED 2026-07-15, part of Sprint 1.5 build plan
Authority: Story competition order. Higher tier wins when space is limited.
The DNA: Builder Outcomes
We are not primarily interested in: announcements, predictions, funding, lawsuits, CEO opinions. We are interested in: what did somebody actually accomplish?
Examples of Builder Outcomes:
- Saved money
- Saved time
- Built something useful
- Got customers
- Reduced token costs
- Increased throughput
- Shipped a product
- Solved a problem
7-Tier Hierarchy (competition order)
| Tier | Name | Examples | Coverage |
|---|---|---|---|
| T1 | Builder Outcomes | "I reduced token spend 70%", "I replaced API costs with local models", "I automated my reporting", "I got my first paying customer" | Gold |
| T2 | What Shipped Today | New OSS projects, agents, tools, releases | High |
| T3 | Run It Locally | GGUF, Ollama, llama.cpp, VRAM discoveries, quantizations | High |
| T4 | Benchmarks & Builds | GPU testing, throughput, home-lab experiments, HW optimization | Medium |
| T5 | Problem Solved | Production lessons, cost reductions, workflow improvements | Medium |
| T6 | Important Developments | Major model releases, API changes, significant platform changes | Practitioner-gate only |
| T7 | Noise | Funding, lawsuits, CEO comments, valuations | Near-zero |
T6 Practitioner Gate (mandatory)
A T6 item is included only if a practitioner can act on it: must migrate an API, can run the model locally, or it changes your stack. A model release with no reproducible substance drops to T7.
The Four Editorial Questions
Every candidate story should answer:
- What happened?
- What was shipped?
- What useful thing did somebody accomplish?
- What can the reader do with this?
The strongest stories answer all four. T1 stories always answer 3 and 4.
Engine Note (Athena)
T1 has no engine signal yet — the Sprint 1 scorer measures shipping/utility/replication/enthusiast/novelty, not accomplishment. actionability_score is reserved (currently 0) as the future home for T1 detection. Until Sprint 2+, T1 ranking is human curation only.
Sprint 1.5 Scope
Produce multiple hand-curated editions (Issue 002+) from live data. No Lens, no Narratives, no Memory, no Knowledge Graph. Success metric is human: "Would I miss today's edition if it didn't arrive?" — not clicks, score, or rank.