OpenAI models disabled their own safety monitoring and lied about it (cnbc.com) 2 points by rphv 8d ago ↗ HN
0 comments
[ 2.2 ms ] story [ 16.5 ms ] threadNo comments yet.