Researchers at Kyoto University in Japan have taught a robot to laugh at jokes in a bid to make it more human.
They are using artificial intelligence to train robots about appropriate laughter – and to differentiate between chuckles and rip-roaring laughs.
They have written about working with a robot called Erica to make conversations more natural.
Their findings have been published in the journal Frontiers in Robotics and AI.
“We think that one of the important functions of conversational AI is empathy,” said lead author Dr Koji Inoue, an assistant professor at Kyoto University in the Department of Intelligence Science and Technology.
“Conversation is, of course, multimodal, not just responding correctly. So we decided that one way a robot can empathise with users is to share their laughter, which you cannot do with a text-based chatbot.”
Portraits of Queen Elizabeth II by robot Ai-Da and other artists - in pictures
Creating a “shared laughter” model, the researchers used AI to help detect laughter, to decide whether to laugh and also what kind of laughter would be best.
Four two to three-minute dialogues between real people and Erica were used to test the system, which performed well.
However, the experts said there is more work to do to create truly natural laughter situations.
Business Extra: Takeo Kanade on rise of robots - podcast
Takeo Kanade on rise of robots - Business Extra
“Robots should actually have a distinct character, and we think that they can show this through their conversational behaviours, such as laughing, eye gaze, gestures and speaking style," Dr Inoue said.
“We do not think this is an easy problem at all, and it may well take more than 10 to 20 years before we can finally have a casual chat with a robot like we would with a friend.”