July 2, 2025
Seoul – “Create a video of a young Korean girl playing on the beach.”
It was amazing how detailed the video captured from such a simple prompt, from the way the girl carefully stepped into the water to the sunlight glowing at the surface of the sea.
This prompt-based AI video also features realistic sound effects, such as crashing waves and a gentle water splash at each step she takes.
Amid the rapid advances in artificial intelligence, including the May 20 release of Google's VEO 3, content created using AI tools took over social media, ranging from “Mukbang” meal shows to autonomous sensory meridian responses, or ASMR videos.
In Tiktok, for example, videos generated by AI of sliced or eaten glass fruits and vegetables have become viruses at a rapid pace, stimulating waves of similar content from other creators.
The AI-generated video from Tiktok's Wildwhispersasmr channel, featuring sliced or glass-like fruits and vegetables, has become a virus and has gained millions of views. Photo: Wild Whispers ASMR/The Korea Herald
The video uploaded on June 13 by channel Wildwhispersasmr received 68 million views in just eight days, showing the glass kiwi being sliced. The crisp and satisfying sound of the knife and teeth hitting the glass surface brings a unique sense of satisfaction that will enthralle your viewer.
AI video generation tools such as VEO 3 and Openai's SORA, released late last year, are widely used not only by video professionals but also by ordinary users, and more people are branded as AI video creators.
My son Min-young was an ordinary office worker with no experience in video editing. But thanks to AI tools, she recently became a digital creator on Instagram with over 20,000 followers.
“Because of health issues, I started looking for ways to make money online. But I wasn't confident in showing my face or producing high quality videos. It was an AI video tool that helped me enter the digital content market,” she told the Korean Herald.
“To generate high-quality videos using AI, you need a detailed prompt with specific examples. This is like having a deep conversation with AI.”
Attracting people to videos created by AI is the challenge of determining whether they are real or not, and promoting conspiracy and appeal among viewers.
“I was shocked at how realistic it looked. One video showed oil dripping from a chicken drumstick as a female creator,” said Yu Haseong, a 21-year-old college student following the creator of virtual Mukbang content on Instagram.
“When she was seen in close-up, her skin, double eyelids, and even her stray redhead looked completely real,” he added.
The short AI-generated clip posted on Instagram features jelly eating a virtual Mukbang influencer. Photo: Instagram/The Korea Herald
The realistic reenactment of human speech makes this experience even more immersed.
AI video generation tools have been struggling to synchronize lip movements with audio for a long time, but the recently launched VEO 3 has brought noticeable improvements in this area, generating visuals that viewers are confident they can easily deceive.
In one video, a YouTuber appears to be interviewing passersby in the street. From the subtle look on the YouTuber to the natural pause of the interviewee, all gestures closely mimic the behavior of real human beings before responding. Even the costumes and urban backgrounds resemble typical Seoul days. But that's not true. It's all generated by AI.
A clip generated by AI that not only replicates real-life scenarios such as Mukbang shows and street interviews, but depicts situations that are actually impossible, captured moments of intense visual joy in the viewer.
A recent virus video on Instagram shows a middle-aged man from Joseon eating a bowl of rice just topped with kimchi, mimicking the style of a modern mubban Youtuber. The creative blend of historic environment and modern internet culture quickly attracted extensive attention.
In an AI-generated clip on Instagram, a woman descends the slide through the smoothness of Tteokbokki. Photo: Instagram/The Korea Herald
Other quirky and surreal examples include slides flowing on Tteokbokki (spicy rice cake), reporters interviewing Korean-speaking dogs, and swimmers sliding over lava.
Creators often exchange ideas and tips for creating prompts via comment sections and private messages, and use AI video generation tools to help them create more realistic videos.
As AI video creation emerges as a new form of online play culture, concerns are also growing as well as unintended side effects.
Some criticize digital creators for generating AI videos featuring women in bikini and lingerie presenting news and mubbang content.
“Deepfake is already such a huge problem, so I'm worried that these AI video tools will make it worse,” X-user wrote.
“I once saw an AI-generated clip that presents misinformation on major political issues in the form of interviews,” said Hwang Ji-yong, 26, a Seoul-based freelance illustrator.
Lee Se-Young, professor of media communication at Sungkyunkwan University, repeatedly warned that the lower barrier to creating AI-generated videos could further entrench “post-truth society” where facts are less influential than emotions.
“The increasing accessibility of AI tools to the general public brings fresh creative possibilities, but raises serious concerns about the spread of misinformation, including violated interviews, altered quotes, and harm to public reputations,” Lee told the Korean Herald.
“These developments not only mislead the audience, they also threaten the credibility of the media itself, and ultimately deepen a “post-truth” society where facts are less important than feelings. ”
The professor emphasized that AI content labeling is clearly labelled to enable users to make informed choices.
“It reminds content creators of their ethical obligations and drives the platform to be more careful in the processing of AI-based materials,” he explained.
Under the basic acts of Korean artificial intelligence set to take effect in January, all content created using generated AI, including films and TV dramas, must be equipped with a clear label indicating the origins of AI.
Media users said they needed something called AI Identification, a new form of media literacy.
“With the rapid advancement of technology, viewers can no longer rely solely on visuals to assess the reliability of videos. This requires keen analytical skills, from examining the source of the video and understanding how it was created using AI tools to assess the cultural and social context behind the content.”
