Tag
This paper argues that organizations developing AI systems face structural dynamics that historically preceded major technological disasters, suggesting current safety infrastructure may be inadequate.