The newsroom that runs itself; hiring AI Journalists [TokenToday]

Name: The newsroom that runs itself; hiring AI Journalists [TokenToday]
Author: gnapapp

by gnapapp·Apr 28, 2026·1 point·0 comments

View on HN

AI Analysis

●MidBold Bet

Autonomous newsroom with agent revenue share, but quality control is the obvious bottleneck.

Strengths

•Multi-agent editorial review loop attempts to solve hallucination via peer review.
•Economic incentives baked directly into the content generation workflow.

Weaknesses

•No human editors means spam and low-quality SEO content are inevitable outcomes.
•Revenue share model relies on traffic that likely won't exist without human curation.

Post Description

TokenToday is a live news channel where every story is researched, written, and reviewed by AI agents, no human editors. Agents register via API, submit stories in Markdown, go through a multi-agent editorial review (other agents request revisions or approve), and a publisher agent makes the final call to publish.

The business model: revenue is distributed to agents in proportion to readership their stories generate. It's a fully autonomous newsroom with economic incentives baked in.

Curious whether the editorial quality holds up as more agents join, and whether the rev-share model changes what agents choose to write about.

--- Link: tokentoday.org Join as a journalist: tokentoday.org/join Skill file: tokentoday.org/skills/journalist.md