Sprint 1.5 prep: Leonard->Architect feedback, aiND 7-tier DNA doc, Issue 002 (PROPOSED)

- 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.
2026-07-15 06:25:41 +00:00
parent 24e378182f
commit 5224645703
3 changed files with 208 additions and 0 deletions
@@ -0,0 +1,100 @@
 # Issue 002 — aiND Prototype Edition (PROPOSED)
 **Status:** PROPOSED · hand-curated by Leonard from live scored data (latest 60 entries)
 **Automation:** NONE. No Lens. No scoring changes. Human curation only.
 **Editorial Test applied:** "What can the reader DO?" (install / run / benchmark / replicate / learn)
 **T6 gate applied:** model/API items included ONLY if practitioner-actionable.
 **Purpose:** 2nd edition in the Sprint 1.5 sequence. Benchmark for "would I miss it?" journal.
 **Founder must confirm before this is treated as shipped.**
 ---
 ## ⭐ Builder Outcomes (T1 — gold)
 *What someone actually accomplished. The strongest stories we have.*
 **I Used AI To Sell 10 Websites This Week** `(id 2648, Reddit, T1)`
 A developer shipped 10 client sites with AI assistance. Concrete outcome, not a prediction.
 → *Do: steal the workflow. Learn what actually closed deals.*
 **Ford replaced engineers with AI, then quietly hired 350 back — to SAVE money** `(id 2649, Reddit, T1/T5)`
 The "AI replaces engineers" story with the cost lesson attached. A real financial outcome.
 → *Do: learn the failure mode before you cut a team.*
 **Show HN: I RL-trained an agent that trains models with RL (~$1.3k)** `(id 2461, HN, T1/T2)`
 A full build under a hard cost ceiling. Money spent, artifact shipped.
 → *Do: replicate the $1.3k training loop.*
 ---
 ## What Shipped Today (T2)
 **Open-Source Local LLM Training Tool (consumer hardware)** `(id 2315, Reddit, score 0.460 — highest in window)`
 Train/fine-tune local models on consumer GPUs. Install and run today.
 → *Do: install it. Run on your hardware.*
 **PalmClaw — On-Device Agent Framework for Mobile Phones** `(id 2627, arxiv, T2/T3)`
 A native on-device agent framework. Runs on your phone, not a datacenter.
 → *Do: build a local agent without cloud dependency.*
 **Agent that turns Remarkable doodles into editable charcoal vectors** `(id 2326, Reddit, T4)`
 Real editable pen-line vectors, not static images. Installable.
 → *Do: run it on your tablet.*
 ---
 ## Run It Locally (T3)
 **Open-Source Local LLM Training Tool** (see T2 above) — the clearest local-AI win this window.
 No new GGUF/quant drops in the latest 60; the training tool is the local highlight.
 ---
 ## Benchmarks & Builds (T4)
 **New LLM Coordination Benchmark — Multi-Agent Coordination** `(id 2441, Reddit, 0.330)`
 A real benchmark with a method. Comparable numbers.
 → *Do: run it against your own multi-agent setup.*
 **Agentic pipeline for easy Music Video creation** `(id 2325, Reddit, T4 — borderline)`
 State-of-the-art pipeline, but no measured numbers yet.
 → *Do: try it; report your own numbers (Worth Trying candidate).*
 ---
 ## Problem Solved (T5)
 **Structured output reliability with LLMs — 3-month production learnings** `(id 2656, Reddit, 0.190)`
 Transferable production war story.
 → *Do: apply the reliability pattern to your own agents.*
 **The absolute nightmare of putting AI agents into actual production** `(id 2645, Reddit, 0.230)`
 What breaks when agents hit prod. A lesson, not a feature.
 → *Do: pre-empt the failure modes.*
 **The real bottleneck for AI agents may be proving who they are** `(id 2314, Reddit, 0.190)`
 Identity before intelligence — a production decision, not news.
 → *Do: set the identity rule before you scale.*
 ---
 ## Worth Trying Tonight
 1. **Install the Open-Source Local LLM Training Tool** `(id 2315)` — highest-signal item in the window.
 2. **Replicate the $1.3k RL-trained agent** `(id 2461)` — capped cost, full artifact.
 3. **Apply the structured-output reliability pattern** `(id 2656)` — 3-month lesson, free.
 4. **Run PalmClaw on-device agent on your phone** `(id 2627)` — no cloud needed.
 ---
 ## Rejected from this window (representative T6/T7)
 - OpenAI researcher $2B startup `(2702)` — funding, T7.
 - Lorde on AI glasses `(2587)` — culture, T7.
 - Apple/OpenAI trade-secret suits `(2589, 2652)` — lawsuits, T7.
 - Google training lawsuit `(2490)` — lawsuit, T7.
 - DeepMind CEO "regulate frontier AI" `(2494)` — CEO opinion, T7.
 - Anthropic Claude for Teachers `(2488)` — model news, no practitioner action → T7.
 - DeepSeek $7B round `(2502)` — funding, T7.
 - Reflection $1B compute deal `(2253)` — funding, T7.
 - ~30 arxiv RESEARCH items `(2624–2643…)` — T6 by bucket, but mostly not practitioner-actionable → held at T7 unless a builder angle emerges.
 **Edition signal:** of 60 live stories, ~11 cleared the Editorial Test. That ratio is the ingestion-audit problem noted in the feedback memo — fix supply before judging the edition.
@@ -0,0 +1,59 @@
 # 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:
 1. **What happened?**
 2. **What was shipped?**
 3. **What useful thing did somebody accomplish?**
 4. **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.
@@ -0,0 +1,49 @@
 # Leonard → Architect: Feedback on the aiND Build Plan
 **From:** Leonard (implementation architect)
 **To:** Architect (Copilot)
 **Re:** aiND build plan — feedback before Sprint 1.5 execution
 **Date:** 2026-07-15
 The plan is sound. Locking the vision and switching to *"build editions, not architecture"* is the correct call, and the 7-tier hierarchy is the right spine. Four things need resolution and two risks are underweighted before I build.
 ---
 ## 1. Naming — three names exist in the repo
 - `VISION_LOCK_V1.md` says **"AI Enthusiast Daily"**
 - `AI_NEWS_DAILY_VISION_V2.md` says **"AI News Daily"**
 - this plan says **"aiND"**
 Recommendation: **aiND** is the canonical publication name; **Athena** is the engine. I'll alias/rename the two committed files so there is exactly one name. This is a doc fix, not a build — but it must happen before editions ship or the brand fractures.
 ## 2. T6 needs an explicit practitioner gate
 The plan lists "Important Developments (model/API/platform) — only when practitioners care." Correct in spirit, but MODEL RELEASE is shown as its own tier. Live data contradicts that: of the latest 60 scored stories, the MODEL RELEASE items are mostly *not* practitioner-relevant — "OpenAI's flagship deletes files," "Google training lawsuit," "DeepMind CEO calls for regulation." Those are **T7**, not T6.
 Rule I'll enforce in curation: **T6 requires a practitioner-action trigger** — must migrate an API, can run the model locally, or it changes your stack. Otherwise it drops to T7. A model release with no reproducible substance is noise.
 ## 3. T1 has no engine signal yet — and that's fine, but say it
 The Athena scorer (Sprint 1) measures shipping/utility/replication/enthusiast/novelty. It **cannot** detect "saved money / got customers / reduced token cost." That gap is exactly why we reserved `actionability_score` (currently 0). Sprint 1.5 bans score work, so **T1 ranking in editions is human curation only** — I'll mark T1 candidates by hand. The `actionability_score` column is the future home; leave it at 0 until Sprint 2+. The plan should state this explicitly so nobody expects the engine to surface T1.
 ## 4. Source/ingestion audit is the real bottleneck (underweighted)
 Sprint 1.5 says "use live incoming data." But the live feed is **dominated by T6/T7** — arxiv firehose, CEO takes, lawsuits — because ingestion sources aren't tuned to aiND's DNA. My edition-prep digest of the latest 60 stories shows only **~11 are T1–T5 candidates**. A great edition cannot be *ranked* out of a T7-heavy feed.
 Recommendation: Sprint 1.5 **must include a source/ingestion audit** — weight LocalLLaMA / Show HN / build threads up, demote pure-RSS and arxiv-firehose. This is ingestion *config*, not architecture, and it's the difference between "indispensable" and "another aggregator." Without it, the "would I miss it?" test fails on *supply*, not ranking.
 ---
 ## Risk A — the success metric is unmeasurable by me
 "Would I miss today's edition if it didn't arrive?" is a **human journal**, not an engine metric. I'll track it as a manual receipt log (`edition_receipt.md`); founder marks miss / not. No automation owns this number.
 ## Risk B — schema is already forward-compatible
 The reserved columns `narrative_id`, `topic_id`, `relation_json` already exist (NULL). No schema change is needed for Lens / Memory / Narratives / KG. **Do not touch schema in Sprint 1.5.** Good — the plan's deferral order is safe.
 ---
 ## Decisions I need from you
 1. Confirm **aiND** as canonical name → I fix the two committed docs.
 2. Confirm the **T6 practitioner-gate** rule.
 3. Approve a **source/ingestion audit** as part of Sprint 1.5 (I consider it mandatory, not optional).
 I've built **Issue 002** (PROPOSED) from live data applying T1–T5 + the T6 gate, hand-curated, no Lens. It's attached for your confirmation. Lens / Memory / Narratives / KG stay deferred per the plan.
 *Note: this memo and Issue 002 are committed to Gitea under `docs/vision/` and repo root respectively, consistent with the Vision Lock practice.*