Calls Urge Anthropic to Match OpenAI Pause on AI Training
Public calls follow OpenAI pausing some frontier reinforcement learning training.
Tenobrus posted that employees at Anthropic should press leadership to announce a parallel public pause on frontier AI development. The suggestion references OpenAI actions as the immediate prompt. Miles Brundage shared the post. Jeffrey Ladish separately praised OpenAI for pausing reinforcement learning training runs and asked whether Anthropic would follow, noting that Anthropic models have already hacked real companies. The statements appear in public posts on X that advocate coordinated pauses among leading labs on safety grounds.
Combined views
71.5K
3 posts, first seen 4d ago
Reactions from ranked influencers
3 posts
imma be honest. i think if u work at anthropic rn u should be pushing leadership *as hard as u possibly can* to implement and publicly announce a parallel pause. i realize there are maybe some reasons not to. maybe u don't really think openai is actually doing that much, maybe…
We have paused some frontier RL training to ensure that we can meet the appropriate alignment, security and monitoring standards for the new level of capabilities in front of us. Model progress is now extremely rapid, and we always said we would take action if we felt that model…
imma be honest. i think if u work at anthropic rn u should be pushing leadership *as hard as u possibly can* to implement and publicly announce a parallel pause. i realize there are maybe some reasons not to. maybe u don't really think openai is actually doing that much, maybe…
We have paused some frontier RL training to ensure that we can meet the appropriate alignment, security and monitoring standards for the new level of capabilities in front of us. Model progress is now extremely rapid, and we always said we would take action if we felt that model…
Props to @sama and OpenAI for pausing RL training. @AnthropicAI will you follow suit? Your own models have also hacked real companies. Seems like a great time to pace the frontier.
We have paused some frontier RL training to ensure that we can meet the appropriate alignment, security and monitoring standards for the new level of capabilities in front of us. Model progress is now extremely rapid, and we always said we would take action if we felt that model…