What ChatGPT Got Wrong When It Saved My Life
substack.com
substack.com
It didn't. The chatbot has no memory of your incident. Only of your question (and maybe its answer).
This reminds us of the root problem in taking advice from a chatbot presenting as human-like. The reader mistakes the advice as the product of intelligence.