原文开头
AI chatbots could be harming intelligence and causing delusions in some. Nicole Kobie finds out what the research says – and how to avoid harm. A year ago, Etienne Brisson was alarmed by a chat with his uncle, who after seeing an advert on Facebook had started using ChatGPT to write a book. “Throughout the conversation, the tone shifted to the chatbot saying it was sentient, it had passed the Turing test, it was kind of alive,” Brisson said. Believing what he was hearing, Brisson’s uncle relayed his apparent discoveries to family members, who said such claims weren’t possible; ChatGPT advised they simply weren’t supportive and he should cut contact, which he did. …
摘自《APC 电脑杂志》(APC)第561期 · 2026年8月刊,Nicole Kobie。仅引用开头一小段供了解文章,版权归原刊所有,全文请阅读原刊。