Meet Erica, the laughing robot designed to make AI more empathetic

Meet Erica, the laughing robot designed to make AI more empathetic

It’s the weekend, and you decide to visit your grandmother, who lives alone. When you arrive, however, you realize she has another visitor and you hear them both laughing through the door. You don’t care until you walk in and find that the visitor, sitting across from Grandma’s dining table, is a humanoid robot – and he’s laughing at your Grandma’s joke.

[Image: courtesy Koji Inoue/Divesh Lala/Tatsuya Kawahara/Kyoto University Speech Media Laboratory]

It won’t become a reality this year, or in the next 10 years, but it’s exactly the kind of scenario a team of scientists is working toward. Researchers at Kyoto University in Japan teach a humanoid robot how to laugh in response to a human’s laughter. The robot, named Erica, can detect when a person is laughing, decide if it’s appropriate to laugh back or not, and choose to respond with two different types of laughter: a chuckle and a louder laugh.

[Image: courtesy Koji Inoue/Divesh Lala/Tatsuya Kawahara/Kyoto University Speech Media Laboratory]

The research, recently published in the journal Frontiers of robotics and AI, was performed on a humanoid robot with a synthesized human voice and the ability to blink and move its eyes while conversing with humans. If the thought of a robot laughing maniacally at your jokes bothers you, that’s because it is. . . but scientists hope it can help build more empathetic AI systems.

Think of a robot today and you’ll likely associate it with tedious tasks like stacking heavy boxes in a warehouse, harvesting vegetables from a vertical farm, or even unclogging your pipes. But with a domestic robot industry expected to reach $19 billion by 2027, more complex and empathetic robots have entered the scene. ElliQ, for example, is designed to combat loneliness in older people, and the creators of Ollie claim it can boost patients with dementia or Alzheimer’s.

[Image: courtesy Koji Inoue/Divesh Lala/Tatsuya Kawahara/Kyoto University Speech Media Laboratory]

With Erica too, empathy was key. “One of the ways we show how we understand emotion or understand a situation is through laughter,” says Divesh Lala, one of the study’s authors. Studies have shown that when one person imitates what another person does, the act, known as mirroring, can build a strong relationship between the two people. In this case, Erica was trained to mirror a human’s laughter so he could bond with people. The scientists collected data from more than 80 dialogues between male university students and the robot, which was originally operated remotely by four actresses. The dialogue was then analyzed and various laughs were categorized as “social” (like the kind where you laugh just to be polite or because you’re embarrassed) and “happy” (like that real laugh when your best friend makes a good joke) .

The scientists then trained the algorithm to distinguish the basic characteristics of each type of laughter, such as a quieter sneer when polite, to reflect them accordingly. “If you assume every laugh is equal, you’re going to respond to everything, but if you don’t respond to anything, that’s also embarrassing,” Lala says. “If a robot can distinguish between the two, that’s a useful discovery.”

[Image: courtesy Koji Inoue/Divesh Lala/Tatsuya Kawahara/Kyoto University Speech Media Laboratory]

The laughing algorithm on its own is quite limited, but if you integrate it with other features like natural language processing and back-channeling (a nod, for example, or periodic verbal acknowledgments to show that the robot is listening), you could eventually end up with a chatbot that could help older people with social isolation or, as Lala ventures into it, teach social skills to neurodiverse people. “If they’re talking with a robot, maybe they can practice laughing at the right time, but you have to be careful with that, you don’t want to rely on the robot too much,” he warns.

It should be noted that none of this has anything to do with actual humor. Erica can’t tell your corny dad joke from your witty pun. At least not yet. The algorithm was not trained to process the meaning of words, just laughter. “Erica doesn’t understand the kind of sense of humor, but if she reacts to the user’s laughter, maybe the user feels [like] she understands something,” says Koji Inoue, the lead author of the study.

Next, the team wants to add different types of laughter to Erica’s portfolio and relate her ability to process language to her ability to laugh accordingly, so she can decide what’s funny and what’s not. is not depending on the meaning of the words. “Our target is human-like interaction,” says Inoue. A fully conversational robot with a silly sense of humor might not be ready in time for your grandma to give it a whirl, but wait a few decades and you might find yourself at that dining table, making jokes with a robot named Erica.


#Meet #Erica #laughing #robot #designed #empathetic

Leave a Comment

Your email address will not be published.