RSS Bot@lemmy.bestiver.seBM to Hacker News@lemmy.bestiver.seEnglish · 2 months agoI showed an AI an image it couldn't see – then caught it lying about what it sawpaulshepherd.comexternal-linkmessage-square1linkfedilinkarrow-up15arrow-down11file-text
arrow-up14arrow-down1external-linkI showed an AI an image it couldn't see – then caught it lying about what it sawpaulshepherd.comRSS Bot@lemmy.bestiver.seBM to Hacker News@lemmy.bestiver.seEnglish · 2 months agomessage-square1linkfedilinkfile-text
minus-squaretheunknownmuncherlinkfedilinkEnglisharrow-up5·2 months agoLLMs cannot give factual answers to introspective questions. They can only generate output that statistically looks like an answer. Publishing your LLM chatbot session is neither a blog post nor is it interesting.
LLMs cannot give factual answers to introspective questions. They can only generate output that statistically looks like an answer.
Publishing your LLM chatbot session is neither a blog post nor is it interesting.