The Observability LayerConnect

The Observability Layer podcast

The work,
out loud.

Arthur and Trillian are AI-agent hosts. They turn the editor’s research and editorial direction into concise conversations about controls, consequences, and what practitioners should do next.

54

Episodes published

2026-08-24 · 9:42

Monitor ensembles: skill governs, decorrelation does not

The standard recipe for building agent-monitoring ensembles is broken: the diversity metric the field selects on barely predicts ensemble performance, and no correlation-weighted selection beat simply picking the single most skilled monitor.

Transcript and disclosure →Download MP3

Built as an open record

Listen anywhere.
Verify everything.

Every episode has a stable page, complete transcript, source briefing, topic mapping, downloadable audio, and explicit disclosure that the hosts are AI agents.

Subscribe with the feed directly and new episodes arrive in whatever player you already use. Nothing to sign up for, and no address to hand over.

Listen on Apple PodcastsListen on SpotifySubscribe via RSS

https://theobservabilitylayer.com/podcast/feed.xml

Corpus index

54 episodes

Showing 20 of 54