#ai-uncertainty-communication

[ follow ]
#language-model-bias #probability-interpretation #human-ai-alignment #linguistic-variation
Artificial intelligence
fromFortune
1 day ago

We studied chatbots and language and saw a huge problem: They mean 80% when they say 'likely' but humans hear 65% | Fortune

Large language models fail to align with humans when communicating uncertainty, using probability words like 'likely' and 'probably' to represent different numerical values than humans expect.
[ Load more ]