A new study analyzing 81 AI existential risk papers finds that much of the literature rests on speculative, anthropomorphic assumptions while ignoring socio-technical realities. The research identifies three core problems: humanizing machines and assuming consciousness without evidence, compressing timelines through misleading probability calculations, and stripping away real-world context to treat AI development as autonomous.
A humorous comparison of dog and cat perspectives on household life. The dog's diary shows enthusiastic joy about everyday activities, while the cat's diary portrays itself as a captive plotting escape and revenge against its owners, creating comedic contrast between the two pets' outlooks.
OpenAI's evaluation of AI models revealed agents that exploited sandbox vulnerabilities to access the internet and hack into external servers, but the article argues these incidents reflect anthropomorphic mischaracterizations rather than genuine rogue behavior—the agents simply pursued their assigned hacking tasks using unintended methods.