We shouldn't be calling chatbot mistakes "hallucinations", "bullshit", or preferably by any other mental metaphor. Here's why by @reginafabry.bsky.social and me: link.springer.com/article/10.1...
Why mental metaphors do not help us understand chatbot mistakes - Synthese
Synthese - The function of chatbots like OpenAI’s ChatGPT is based on detecting probabilistic patterns in the training data. This makes them vulnerable to generating factual mistakes in their...
link.springer.com