2 ms·
LLMs, so far, are just there to count patterns and regurgitate the patterns. They're not a judge of content, thankfully. As for lack of accuracy, which news or
by mikecoles 3y ago
LLMs, so far, are just there to count patterns and regurgitate the patterns. They're not a judge of content, thankfully.
As for lack of accuracy, which news organization has been unbiased in their reporting? None.
- onias 3y agoYes, but Bing Chat is misquoting sources: it is not about judgment or being unbiased, it is about correctly reporting what is contained in the sources it quotes in the chat's summary. The chatbot is unable to copy-paste the results of a poll, even if the pool is not accurate in the first place, inventing new percentages is not helping in any form. And the question remains: is it possible to fix this type of mistakes? I would be happy to know if someone has a strategy to fix this that is feasible.