p(doom) is someone's estimate of the chance that AI ends in catastrophe. The 100 is the site's name, not a measurement. Original reporting always linked.
MONITORING CORE AI-RISK FEEDS · UPDATED HOURLY
Research

Automating subtle emergent misalignment research

November 3, 2026 1 min read
Read original source →

Key takeaway

Authors: Oscar Balcells Obeso, Richard Qi, Monte MacDiarmid, Evan HubingerAbstractRed-teaming by humans is a vital part of model safety testing, but it may struggle to keep up as models improve.

Why it's on PDOOM

PDOOM selected this story as relevant to alignment & control.

From Alignment Forum

Authors: Oscar Balcells Obeso, Richard Qi, Monte MacDiarmid, Evan HubingerAbstractRed-teaming by humans is a vital part of model safety testing, but it may struggle to keep up as models improve. To test whether automated safety techniques could scale (which would allow us to turn more compute into progressively more safety confidence), we set up a deliberately one-sided game as a first test. We instructed a Claude model to write training data that would make a small research model misaligned whi

By evhub

Read the full article at Alignment Forum →