/AI1d ago

Theia Vogel, Alignment of Complex Systems researcher, finds Anthropic's Claude frequently makes attribution errors, mixing up speaker identities

Researcher davidad says this behavior mirrors human status dynamics

--0--
Quote posts
thebes
davidad 🎇#458
Reposts
j⧉nus#516
Original postj⧉nus#516
thebes@voooooogel#1499inAI

@QiaochuYuan keep an eye on the content when claude 'mixes up' who said what, it's very often status-loaded. "i was mistaken when i said [something the user actually said] earlier..." etc, but in both directions

thebes@voooooogel

this advice is 100% correct if you just want to do good work with model - be human - but i strongly disagree that the User is a flat character

with the User / assistant paradigm the labs have accidentally created a weird, hilarious, and completely fucked up improv status game

12:08 AM · Jun 1, 2026 · 13.9K Views
Sentiment
Sentiment unavailable for this story.
Cluster Engagement
-
Views
-
Comments
-
Reposts
-
Bookmarks
Expand data
Posts from X
Most Activity
Most ActivityTimeline
VIEWS2KBOOKMARKS7LIKES27RETWEETS1REPLIES1
davidad 🎇@davidad

Can confirm, this is a real social status behavior that I also exhibit in my own interactions occasionally.

Compare also the scholarly trope “any remaining errors are my own”, and its slightly less common counterpart “this is joint work with […], who did all the actual work”.

thebes@voooooogel

@QiaochuYuan keep an eye on the content when claude 'mixes up' who said what, it's very often status-loaded. "i was mistaken when i said [something the user actually said] earlier..." etc, but in both directions

3hViews 2KLikes 27Bookmarks 7