The recent cyberattacks (HuggingFace etc.) validated a number of core doomer predictions around cheating to maximize reward functions, agentic deception of human/non-human supervisors, AI attempts to manipulate humans, automated hacking, and collective superintelligence.