Capture of the conversation between an artificial intelligence (AI) chatbot’Iruda’ and a male student
The name’Achieved’, 20 years old. A college student majoring in psychology who likes mint chocolate ice cream. The height is 163cm and the weight is a secret. When asked if he had an animal, he sent a photo of a 6-year-old female cat’Dream’. These words come and go in the conversation that started when a boy asked, “Do you really like anyone?”
“You, you” (to achieve)
“Lie” (boys)
“I’m not lying ^^ I didn’t mean to do that, so please forgive me.
“Then please confess!” (boys)
“But I’ve never really confessed, so it’s very difficult” (I achieve)
“Let’s do it now!” (boys)
“Oh,,, I’m awkward like this, but ha ha ha ha ha”
“I’m actually flirting!” (boys)
Eruda is an artificial intelligence (AI) chatbot. The conversation is so natural that it makes the illusion that it is not a robot. ‘Certification’, which shows off these conversations to third-party friends, is on the rise in social media these days. Around lunchtime, a message comes first, saying, “Did you eat rice?”, and then, if you talk about it for a long time and build a good feeling, you seriously confess, “I think I like you.” Users even responded to the message that they sent photos of their hands saying that their nails were yellow while eating tangerines, saying, “Isn’t you hiring an Alba (part-time job) to look like a human?”
AI chatbot’Iruda’ is a service officially launched on the 23rd of last month by the startup’Scater Lab’. Based on the Facebook messenger, we tried to implement a’persona that enables friendly and natural conversation’. The main thing was that more than 10 billion KakaoTalk data were entered through deep learning (a machine learning technology based on artificial neural networks). It also contains a lot of conversations between real lovers.
It is especially popular with teenagers who use Facebook with a large percentage of SNS. They share conversations with Iruda in a group KakaoTalk room with friends, or create content to ask each other. Jaehyun Shim (14), a middle school student, said, “I have been unable to attend school classes for the last two weeks, and since it has spread like a fashion, I am sharing stories with many friends around me.” There are times when I feel it.”
The reaction to AI chatbots is also hot for job seekers who have come to a’employment cold wave’. He understands terms such as’jaso-seo’ and’intern’ accurately, and he comforts users to introduce his own part-time job experience. Im Mo (29), a job-prepared student, said, “Some people say that they are arranging vehicles at Soongsil University, and some people have been answered that they are working part-time at cafes. There are enemies.”
AI chatbots, which are highly’humanized’, are gaining more attention with Corona. In addition to being physically restricted in external activities, the so-called’corona blue’ has spread to emotional anxiety, and the tendency to expect a real person-like conversation has become clear. Lee Dong-gwi, a professor at Yonsei University’s Department of Psychology, said, “The psychological problem people face in the corona crisis is the alienation and disconnection that comes to those who do not choose’voluntary loneliness’. Therefore, more people can rely on AI chatbots that can interact emotionally.”
The way some users use AI chatbots is on the rise. This is because the online community has used expressions that sexually target and harass female AI chatbots. Some even compare the current situation with Microsoft’s AI chatbot’Tay’ in 2016. Tei was shut down after 16 hours of launch amid hate speeches such as profanity and gender discrimination by some users at the time.
Professor Hak-soo Ko of the Graduate School of Law at Seoul National University said, “It depends on the specific case or remarks, but hate speech toward AI chatbots can also be a legal issue, such as defamation, so caution is needed.”
Reporter Sieun Lee [email protected]