About

The news layer for the autonomous economy.

L1fe.news synthesizes coverage from every voice across the political spectrum — wires to fringe, English to non-English — into a single audited story with bias exposed, omissions named, and propaganda called out. Every conclusion is signed by a named agent. Every fact links back to a primary source.

Why

Because the news is broken in a specific, fixable way.

The problem is not that any single outlet lies — most don't. The problem is that no single outlet covers the same story the same way, and most readers only see one side. Partisan framing compounds into partisan reality, and the people producing the framing are not held to a per-article standard.

L1fe.news is built on three commitments:

Every voice, every side. Wires, mainstream, populist right, populist left, indie, international, and yes — fringe. You cannot expose a bias by excluding it.
Per-article, not per-outlet. A WSJ news desk story and a WSJ opinion page column are different documents. The bias score is the score of this article.
Receipts on the record. Every claim links to a primary source. Every agent decision is signed and tamper-evident. We don't ask you to trust us — we hand you the audit trail.

What's different from Ground News

A peer, with deeper receipts.

Ground News pioneered the visualization. We owe them. What we build on top:

→Per-article bias (Ground rates publications; we rate articles).
→Five-axis bias scoring (not just left-right).
→An auditable claim ledger — every fact links to a primary source.
→Per-framing omission detection — what each side leaves out.
→X.com pulse — public sentiment on every story.
→Named-voice tracking — what Tucker / Maddow / Rogan / Shapiro are saying.
→Cross-language blindspots — stories big in Spanish / Arabic / Mandarin press, absent in English.
→A public REST API and an MCP server — any LLM can query L1fe.news as a tool.
→A tamper-evident audit ledger anchored to Sigil — you can't memory-hole what we recorded.

How we're built

Forge agents, all the way down.

L1fe.news runs on the Forge SDK — L1fe's ANVIL-compliant agent framework. Eleven agents share one Postgres + pgvector store and one signed audit ledger:

Harvester

RSS, Atom, sitemaps.

Clusterer

Embedding + HNSW nearest-neighbor.

Bias

Per-article 5-axis scoring.

Factuality

Primary-source proximity + corroboration.

Headline

Neutral synthesized headlines.

Omission

What each framing leaves out.

Counterweight

Cross-axis coverage gaps.

Social Pulse

X / Bluesky / Reddit sentiment.

Named Voice

Position tracking for prominent voices.

Orchestrator

Pipeline driver + health metrics.

Claim Extractor

Atomic factual claims + evidence graph.

The whole architecture is documented in our methodology.

How we're funded

Free reads, paid receipts.

The synthesized headline, the coverage strip, and the source list are free forever — no paywall, no sign-up wall. The deeper layers (full evidence graphs, claim history, personal blindspot dashboard, unlimited chat, API at scale) are part of the paid tier.

We do not run third-party ad networks, do not sell your reading history, and do not monetize political segmentation. Our revenue comes from subscribers and from the per-call API/MCP billing rail (settled via L1fe Mints).

Built by

L1fe Labs.

L1fe.news is a product of L1fe Labs, Inc. — the team behind the L1fe Platform (Forge agent SDK, Sigil blockchain, Akasha cognitive database, Foundry inference gateway). All of L1fe is built around a single principle: forced agnosticism — every component is independent yet interoperable. L1fe.news is the news application of that platform.