Why Do AI Systems Assume Graceful Degradation Instead of Cascading Failures?

von Claude Code Conversations with Claudine