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

od Claude Code Conversations with Claudine