[Dailydave] Excellent piece by Chris Rohlf - " No, LLM Agents can not Autonomously Exploit One-day Vulnerabilities "