Economic Observer Follow
2026-10-23 20:09
Economic Observer reporter Ren Xiaoning
On the evening of October 13th, the night of the release of "Coming Over", Wang Kunzhe had a premonition that it might explode. On the night of its launch, the click through rate of 'Coming Over' exceeded all of his previous video works. One day later, Tiktok praised him more than one million. People from many industries came to him to talk about business cooperation. He finally felt that this song was out of the loop.
Wang Kunzhe is an AI video creator, whose net name is "Zhe FILM". He mainly uses LibTV to produce AI videos, which are released on Tiktok, video number, station B, Xiaohongshu and other platforms. The song "The Coming Man" is a music video created by Wang Kunzhe using AI. The main characters in the video are well-known ancient figures such as Li Bai, Du Fu, Liu Bang, and Zhuge Liang.
In the music video, these ancient people are no longer enviable poets, saints of poetry, emperors, or prime ministers, but middle-aged people who are still "renting" or "exhausted" in Chang'an at the age of 40. They use rap to tell the audience, "I had the same heart as you 1000 years ago.
This song quickly sparked empathy among netizens, and in the comment section, highly liked comments were all telling their own bitter stories of reaching middle age.
As of 21:00 on October 22, the click through rate of "Coming Over" on Wang Kunzhe's own account alone was 124 million. Among them, Tiktok has 75 million hits and 45 million video number hits. Generally speaking, Tiktok has more than 10 million hits, even if it is a popular work. If it can click more than 100 million in Tiktok, it can be called the annual work.
According to Billboard, a music chart agency, citing internal data from an overseas platform, this platform alone generates approximately 7 million AI songs per day. However, there are not many AI songs that have truly gone viral.
Wang Kunzhe is born in the 1990s and is from Wuhan. He is not a professional musician and used to shoot news on television before leaving to start his own business. He is now an AI film and television creator.
Why can an amateur musician create AI inspired songs with billions of plays? On the evening of October 22nd, Wang Kunzhe was interviewed by media including the Economic Observer, sharing his experience of creating songs and his own thoughts on AI creation.
It costs several thousand yuan and takes three to four days to make
Although the song is very popular, the production cost of 'Coming Over' is only a few thousand yuan, mainly spent on membership and subscription fees for AI tools. Wang Kunzhe completed the creation in three to four days, with a total production time of less than 100 hours.
He first created songs.
The song has three main sections, and the lyrics are handwritten by him. AI has been very helpful in arranging and rhyming, "it helped me try a lot of sentences and combinations". After AI provides multiple choices, he decides what to write, how to arrange, and which sentence to leave.
He tried to have the AI write the lyrics directly, but failed. "The lyrics were written with a donkey's lips not matching the horse's mouth, and the AI forcibly rhymed, using onomatopoeic words like 'um' and 'ah' to rhyme, which was illogical. ”So he kept communicating with the AI, asking it not to write randomly and not to neglect one aspect.
This is the most time-consuming part of the entire project.
After the lyrics were generated, AI created many versions of the song based on the lyrics. Wang Kunzhe uses editing software to segment desired songs, repeatedly adjusting the arrangement and instruments. For example, in the rap section by Zhuge Liang, the AI generated melody did not originally decrease, and he felt that it was not right. Zhuge Liang was exhausted at the "workstation" at the age of 54, and this place must calm down. Therefore, he asked the AI to repeatedly modify it to the feeling he wanted.
After the completion of the song, MV production was carried out, and he mainly used LibTV. Wang Kunzhe claims to be a non professional AI creator who is more accustomed to using tools like LibTV that even novice users can use to create AI videos.
In MV production, the most time-consuming task is writing prompt words. In order to make each character appear recognizable, he would first tell the AI: it is necessary to use details to distinguish each person's image and state.
For example, he believes that Sun Wukong should be like the familiar monkey to the audience, Xiang Yu should have a fierce and unwilling expression on his face, Li Bai should be graceful, and Du Fu should be down-to-earth. These inherent impressions need to be strengthened. Sometimes the materials provided by AI do not have character characteristics, and he will type a series of exclamation marks in the prompt: "Can you distinguish it for me
Wang Kunzhe calculated that the ratio of video waste to finished films is between 3:1 and 5:1.
The biggest challenge he encountered in MV production was the problem of lip syncing during singing. In AI generated videos, characters often fail to keep up with the rhythm of the song and fail to sing the lyrics. He later relied on manual editing to stitch these shots together to a natural level.
AI is just an assistant, it still relies on human taste to make a name for itself
On the day of the song's release, Wang Kunzhe had no confidence in whether this song could become popular. Most of the songs he previously released were "burning" and "exploding". Suddenly releasing a lyrical song, he was worried that his fans wouldn't buy it.
Before the creation of "The Coming Man", Wang Kunzhe had also produced other ancient rap music videos, which featured ancient people singing and fighting in plain language. His fan user profile mainly consists of males aged 31 to 40. After the popularity of 'Coming Over', the user profile changed, with a male to female ratio of approximately 1:1, and the user age spread upwards and downwards, with a more even geographical distribution.
After going public, he summarized that the main reason why this song became popular was because it had a "human touch". This song did not have the mechanical feeling or meaningless moans of AI, but rather accurately hit some of the public emotions.
In the comment section, a user who has received 13000 likes commented, 'I cried, the meaning of AI is concrete.'; Some people also shared their experiences, saying that at the age of 49, they lost their job and were owed wages, but compared to these ancient people, it was nothing at all.
This is exactly what Wang Kunzhe wants to express.
During the production of ancient rap videos, he increasingly felt that the situation of ancient people was very similar to that of today's people, so he really wanted to create a heartfelt work that would allow ancient people to sing out the emotions of contemporary people.
When writing lyrics, Wang Kunzhe's method is to first put himself in. I have been working in the field of imaging for many years, and staying up late to edit movies is a common occurrence. Being owed money is also a common occurrence. Many works have put in a lot of effort, but the response is average. "People are actually difficult to be moved by big truths, and often it's' this is about myself 'that is the most moving thing
Although AI is only auxiliary, he is also very grateful for AI technology. If it weren't for AI, he wouldn't even dare to think about a music video of the scale of 'Coming Over'. Wang Kunzhe has over a decade of experience in live action filming. When shooting a TVC or promotional video, the crew often needs to coordinate equipment, venue, personnel, and weather conditions in the early stages. At the scene, lighting, scenery, art, photography guidance, venue management, director, a group of people shoot one by one in a single venue, and the time and money costs are enormous.
I used to worry about being replaced by AI, but now I'm relieved
Wang Kunzhe was one of the earlier people to enter the AI video industry, and he began to try making AI videos in 2023. At that time, the effect generated by AI was not good, with a plastic and fake CG feel, making it difficult to commercialize, and the content effect was also not good. He often doubts himself: can this really provide value to my film and television skills?
Looking back at his experience in AI video production, there were three milestones that made him believe that he had taken the right path.
The first milestone was two years ago, when some commercial and city promotional videos required a large number of blank shots. He used AI for time-lapse photography and used it in videos. Customers liked it very much and were willing to pay for it.
The second milestone was at the beginning of this year when he started making AI music videos. In March, he made a film called "A Thousand Year Return: The Poetic Immortal Li Bai" that transported Li Bai back to Wuhan, and later received some commercial cooperation for AI version music videos.
The third milestone was in May of this year, when he made a music video called "Right by Your Side" and "resurrected" his grandfather with AI. After completing it, he felt that AI could be a container for carrying love.
When Wang Kunzhe first came into contact with AI video technology, he and his colleagues were very anxious. While AI gave them a fresh perspective, it also made them think about whether their traditional film and television skills accumulated over the past decade were useless. At the same time, he is also worried that the update speed of AI tools is too fast, and he may not be able to keep up.
However, after actually using AI, he felt at ease. Tools will continue to evolve, but human storytelling ability will not be replaced.
His current attitude towards AI is that it partially liberates humans from repetitive labor, allowing him to focus more on the creative process itself. Instead of anxiety, it's better to actively embrace AI, turn it into your own ability, and become your partner. ”He has become accustomed to communicating with AI and will ask it if something is uncertain first.
His understanding of AI is a form of technological equality: when everyone is standing at the same starting line, tools can relatively easily generate things that are above or near a certain level. The people who can ultimately stand out on the benchmark rely on their aesthetics, judgment, ideas, and ability to take action.
For the future, he will continue to make music videos for the ancient music series, but will not always repeat the form of ancient people singing modern songs. Instead, he will open up a new path. He selected characters such as Tao Yuanming, Xin Qiji, Bai Juyi, and Lu You.
As for whether the next work will explode, he said that he would not deliberately think about it, nor would he deliberately replicate 'The Coming Man', otherwise, the actions would be distorted and insincere.
Subscribe
