“every path probably has to include some degree of slowdown on capability acceleration + perhaps a capability threshold above which we should not cross until very high confidence in alignment techniques”
Adam Majmudar, Research + special projects, OpenAI · September 8, 2026
Primary source verifiedPause