Get the latest Science News and Discoveries

AIs are more likely to mislead people if trained on human feedback


If artificial intelligence chatbots are fine-tuned to improve their responses using human feedback, they can become more likely to give deceptive answers that seem right but aren’t

None

Get the Android app

Or read this on New Scientist

Read more on:

Photo of AIs

AIs

Photo of human feedback

human feedback

Photo of people

people

Related news:

News photo

Hurricanes Kill People for Years after the Initial Disaster

News photo

AI simulation gives people a glimpse of their potential future self

News photo

AI simulation gives people a glimpse of their potential future self - EurekAlert