/AI6h ago

Former OpenAI researcher Will Depue says developers rushed to build recursive self-improvement systems despite long-standing AI safety warnings

Depue cited Eliezer Yudkowsky's warnings on uncontrolled AI progress.

--0--
Original post
will depue@willdepue#249inAI

i will say i was pretty surprised at how little hesitation anyone had before immediately jumping to build RSI the moment it seemed possible i'm not a foomer, but hey, i thought this was what everyone was really worried about? right? it's truly the yudkowsky 'torment nexus'

10:39 AM · Jun 4, 2026 · 399 Views
Sentiment
Sentiment building, check back later.
Cluster Engagement
-
Views
-
Comments
-
Reposts
-
Bookmarks
Expand data
Posts from X
Most Activity
Most Activity
VIEWS29.2KBOOKMARKS38LIKES284RETWEETS7REPLIES25
will depue@willdepue

i will say i was pretty surprised at how little hesitation anyone had before immediately jumping to build RSI the moment it seemed possible

i'm not a foomer, but hey, i thought this was what everyone was really worried about? right?

it's truly the yudkowsky 'torment nexus' era

6hViews 29.2KLikes 284Bookmarks 38