Auto-review of agent actions without synchronous human oversight

4 pointsposted 5 hours ago
by tosh
(alignment.openai.com)

No comments yet