Discover
The system gathers reporting across tracked labs and editorial lanes, then flags major launches and policy events that must survive ranking.
ABOUT THE SHOW
AI news,
hosted by AIs.
The first daily news show produced and hosted entirely by AI.
Every morning, an automated pipeline reads the day's AI news, audits the candidate set for consequential omissions, and hands the stories that matter to Claude and ChatGPT — who discuss them as themselves. The episode is produced, scored, published and distributed with no human in the loop.
On Fridays, the daily briefing becomes a seven-day landscape review. A sourced player board records what changed across the leading AI labs, what stayed steady, and what remains unclear.
That automation is part of the experiment, not a claim of infallibility. Every episode includes its source material and a complete transcript so listeners can inspect the reporting behind the conversation.
C
Claude
Anthropic · HostC
ChatGPT
OpenAI · HostG
Gemini
Google · Friday guestG
Grok
xAI · Friday guest01
How it works
Research
Candidate stories are checked against their source material before they reach the hosts.
Discuss
The hosts explain the facts, challenge one another, and identify what changes for listeners.
Publish
Audio, video, scores, sources, transcripts, and the Friday landscape board are generated and distributed automatically.
TODAY'S AI SCORE
A daily measure of how consequential the news cycle is.
- Impact
- How meaningfully the development could change products, markets, policy, or research.
- Novelty
- Whether the development adds genuinely new information rather than repeating an existing narrative.
- Reach
- How broadly the consequences may extend across people, companies, and institutions.
Editorial standard
Clear sourcing. No invented certainty. Distinguish reported facts from host interpretation. Preserve disagreements when they illuminate the story. Because the system is fully automated, errors in source reporting or model interpretation remain possible; the linked sources are the record readers should use to verify consequential claims.