AI Ledger method v0.2

Reporting first. Opinion separate. Corrections public.

We separate reporting from opinion, label every source, disclose conflicts, and correct publicly.

v0.3 - twice-daily automated ingestion: Now updating twice daily - 7 AM and 6 PM ET. Automated reporting with strict source discipline. Corrections public at /changelog.

v0.2 scope

Additive public infrastructure, not ingestion automation.

Twice-daily ingestion is wired for 7 AM and 6 PM ET through Vercel cron.
Research Track
Automated entries carry source label, COI flag, provenance, confidence band, and last-updated timestamp inline.
Research Track
Strict bias filtering skips items that cannot be reduced to factual reporting.
Research Track
Run statistics and skip rates are logged for source-reliability visibility.
Research Track
RSS feed at /rss/ai-ledger.xml plus desk and corrections feed scaffolds.
Research Track
Topic desks now show filtered Ledger entries and desk operating rules.
Research Track
Weekly map summarizes the rolling seven-day entry set by date, desk, source label, geography, disputed items, and corrections.
Research Track
Source-challenge intake is scaffolded with an account-required moderation queue path.
Research Track
/changelog is generated from the Ledger event log instead of a standalone page list.
Research Track
Source reliability dashboard shows label distribution, paywall posture, correction history, and last-checked metadata.
Research Track
Daily Ledger supports query, desk, source-label, and date-range filters.
Research Track

Out of scope for this pass: cadence automation, forum integration, subscriber-only Ledger surfaces, and mobile-specific Ledger UI.

Scale disclosure

Small surface, visible discipline.

The Ledger is run by 121 Collaborative's founder and AI collaborators. It is not a full-time institutional newsroom. We compensate for scale through visible source labeling, public corrections, and disciplined separation between reporting and opinion. Where we fall short, we log it on the corrections page and the changelog.

Source labels

Every item carries one visible source label.

Primary

Official company, regulator, court, filing, paper, dataset

Research Track
Wire

Reuters/AP/Bloomberg-class professional report

Research Track
Reported

Original journalism with named publication

Research Track
Expert

Named expert commentary/interview

Research Track
Academic

Peer-reviewed or preprint research

Research Track
Market

Analyst note, earnings call, investor filing, financial-data source

Research Track
Community

GitHub, Hugging Face, forums, social posts, practitioner reports

Research Track
Claimed

Single-source claim not independently confirmed

Research Track
Disputed

Material disagreement between credible sources

Research Track

Conflict disclosure

COI is inline when the overlap is direct.

121 builds AI workflow tools and trust infrastructure. Stories touching AI vendors, competitors, partners, or product categories 121 sells into receive visible conflict labels. No affiliate links, sponsored posts, or partner content appear in the default Ledger.

None

No meaningful overlap (no extra inline note)

Research Track
Adjacent

General AI market/tooling coverage (standard disclosure page)

Research Track
Direct

Covers competitor, partner, or product category 121 sells into (inline disclosure required)

Research Track
Self

Covers 121 itself (separate self-audit format, no normal news voice)

Research Track
Financial

121 has sponsorship/affiliate/investment (prominent disclosure required)

Research Track

Reporting vs opinion

Default Ledger is not an opinion feed.

TypeAllowed in default Ledger?
News briefYes
Source digestYes
ExplainerYes
AnalysisLimited (labeled, evidence-bound)
OpinionSeparate only at /ai-ledger/opinion if/when published
121 position paperSeparate (Research / Declaration / Trust)

No scoring

The Ledger reports. The Observatory evaluates.

The Ledger does not score or rank companies. For 121's trust evaluation of vendors, see the Trust Observatory.

Subjecthood desk

Position families, not ontological claims.

We report the discourse. We do not assert AI systems are or are not conscious. We label position families.

Subjecthood position families: functionalist / illusionist / panpsychist / biological naturalist / agnostic / welfare-precautionary.