Hacker News Reader

Interesting idea to increase the scope until the LLM gives suggestions on how to 'hack' itself. Good read!

The escalation of commitment scam, interesting to see it so effective when applied to AI.