Artificial intelligence system laughing empathetically Developed by Kyoto University researchers, it is a step towards building super-intelligent chatbots. Is a future similar to Her coming?

source to laugh where is it from? what is funny At least such a question has confused philosophers and scientists since the time of Plato until today. Earlier, the Greeks believed that this feeling of superiority over others is the source of humor. Sigmund Freud, a German psychologist, also considered humor to be a biological tactic to eliminate repressed desires. And Robin Williams, the American comedian, spent his emotion on the absurdity of this subject in order to make people laugh. But our question remains, what, where is a funny story?! And how can you describe it to a car?

Read more: What is hubris syndrome? Can dictators be reformed?

Empathetic laughing artificial intelligence system

Yes, no one seems to be able to answer the question, “What’s funny?” Be unanimous. Therefore, it is difficult to imagine trying to teach a car dummy to laugh, let alone to understand “laughing”! But a group of researchers from Kyoto University in Japan created a chatbot that sends its signals from Empathetic laughter system receives, are trying to do so. The researchers have described their new technology, which is a kind of laughing framework for the Japanese robot “Erika”, in the journal Frontiers in Robotics and AI.

Don’t get me wrong that car dummies are unable to respond and understand or even laugh at cool jokes. Incidentally, instead, the problem is in developing the subtleties of human humor for the AI ​​system so that routine conversations between robots and humans seem natural.

Dr. Koji Inoue, assistant professor at Kyoto University in the Department of Information Science and Technology, explained:

“We think that one of the most important functions of chatbots Sympathy Is. Of course, the nature of dialogue is multifaceted, and the correct answer is not the only important one; So we found that one of the ways a chatbot can empathize with users is to respond correctly to their laughter, something that cannot be done with a regular text-based chatbot.

A funny thing happened!

At Empathic laughter modelAt first, the human laughs and the artificial intelligence system responds with an empathetic laugh. This approach required the design of three subsystems – one to detect laughter, the second to decide whether to laugh, and the third to select the appropriate type of laughter.

The researchers collected training data by annotating more than 80 dialogues from characters on dating sites, a type of social interaction in which large groups of people interact individually for short periods of time. In this case, a courtship marathon ensued between the students of Kyoto University and Erika, who was fired up from the actresses.

He said this:

“Our biggest challenge in this work is identifying real cases laughing empathetically It was, which isn’t easy because, as you know, most laughs aren’t actually sympathetic at all. “We had to carefully categorize exactly which laughs we could use for our analysis and not just assume that we could respond to every laugh.”

The type of laugh is also important, as in some cases a polite laugh may be more appropriate than a giggle. It should be noted that this experiment was limited to laughter caused by social etiquette as opposed to laughter out of joy.

The empathetic laughing artificial intelligence system is actually a hybrid system

Finally, the group, sense Erika’s humor experimented with the new system of empathic laughing and creating four short dialogues of two to three minutes between a human and Erica. In the first scenario, he would only give polite laughs, followed by happy laughs in the second and third exchanges, and finally a combination of the two.

The team also created two other sets of similar dialogs as base models. In the first, Erica never laughed. In the second, Erica gave a socially polite laugh every time she detected a human laugh, without using the other two subsystems to filter context and response.

The researchers recruited a total of more than 130 people to listen to each scenario in three different conditions – the empathic laughter system, without laughter, and with laughter – and also rated the interactions based on empathy, naturalness, human-likeness, and understanding. , they evaluated. In this evaluation, it was found that the sympathetic laughing system works better than other bases.

The most important result of this research is that we have shown how to combine all three of these tasks in the heart of a chatbot. We believe that this type Hybrid system Laughter is necessary for proper functioning, not merely detecting and reacting to laughter.

Read more: What is the spectator effect? If someone asked for help, would you help them?

Like old friends

There are still many laughing styles that Erica (the robot) can model and train before she’s ready to go into stand-up comedy. Because there are still many other functions and types of laughter to consider, and it’s not an easy task.

He noted that:

“We haven’t even tried to model other types of laughter besides empathic laughter, even though they are more common. However, laughing is only one aspect of the naturalness of conversation.

Ino added:

“Machine humans should actually have a distinct personality so that they can express their communal behaviors such as laughing, looking, gestures and speaking style. So, we don’t think it’s going to be a simple problem, and it’s probably going to be ten or twenty years before we can finally have the kind of intimate conversation we have with an old friend with artificial intelligence.”

Read more:

Your opinion about Empathetic laughing artificial intelligence system What is? Do you think artificial intelligence will be a good friend for us in the future? Share your thoughts in the comments section duplicate Share and technology news Follow with us.

Source:SciTechDaily